Under-representation of older people in clinical trials

Gary H. Mills

Key points

Older people, often with high levels of co-morbidity and polypharmacy, are increasing in number as the population ages.

Study populations in clinical trials have often not mirrored the population that will consume the drugs or treatments being investigated.

Regulators, ethics committees, and funding bodies must recognize the need for inclusion of older people and the complexities this imposes on investigators.

Across all aspects of medicine, funding for research into how best to safely include older people in clinical trials is vital.

Researchers should not be disadvantaged by designing studies to meet the needs of the age range of patients expected to receive the treatment being investigated; this should include provision for older people.

1 Introduction

Older patients consume a large proportion of all medications. In the USA, 88% of those aged 60 used at least one prescription drug, compared to 48% in the 20–59 age bracket. Of those aged 60 and over, 37% used five or more drugs, compared to 8% in the 20–59 age range (1).

As described in Chapter 1, the proportion of older people in the population is rising throughout the developed world. It is therefore imperative from a practical perspective as well as for humanitarian reasons that older people are kept fit and independent for as long as possible.

Part of this process is the provision of suitably developed medication and treatment. Despite this, older people have often been excluded from clinical trials (2), even though in many situations the drugs being tested are used by a population whose age range is higher than the trial population (3). It is vital that drugs are trialed on a population that is similar in age and co-morbidity during the study phase, rather than relying on later reports to determine if complications are reported. This is because drug metabolism is different in older people compared to the younger population. Furthermore, older people are more likely to be taking multiple medications thereby increasing the risk of adverse drug reactions (see Chapter 5).

Randomized controlled studies are the gold standard test of a treatment, but the difficulties in matching populations can cause real difficulty in recruitment. The result is that there is a tendency to look at the simplest possible group, which frequently does not include older people, who are more likely to have co-morbidities. Selecting the best control therapy may also be difficult, as may the selection of primary and secondary outcomes that are most suited to all age groups. For example, in younger surgical cases survival is the key outcome, but in older people the chances of the quality of life being affected adversely is usually much greater. Loss of independence, especially if this involves cognitive problems, would be something that older people might fear most and might justifiably be regarded as the primary outcome. Information on quality-of-life outcomes that might be more commonly affected in older people are vital to allow them to make informed decisions, and vital as well for the doctors trying to select the best forms of treatment. Studies need to be designed and powered with these objectives in mind.

Medical specialists in areas as diverse as neurology and critical care have emphasized the need to include older people in clinical trials and have also stressed that many diseases are more common in older people. They have called upon funders, governments, and ethics committees to examine this problem (4). A study of critical care trials looked at exclusion based on age, gender, and race in 17,000 patients screened for inclusion in a study of acute lung injury. Age over 75 was related to exclusion, some of which could be explained by a high level of co-morbidity. This is important because the average age of the critical care population is increasing, as is the age range for major surgical interventions (5). Older people and women have been under-represented in randomized controlled trials of acute coronary syndromes in the past and this continues to be the case (6). This is particularly important because 60% of deaths related to myocardial infarction occur in patients aged 75 or older (7), and post–myocardial infarction morbidity, including heart failure, are more frequent (8). In fact, one type of heart failure—heart failure with preserved ejection fraction—is a disorder that is largely confined to the over-60s, with a mean age of 75 years (9). In a UK study of incidence of heart failure, the rate increased with age (10), reaching 11.6 new cases per 1,000 in those aged 85 and older. Approximately 80% of all heart failure cases occur in individuals aged 65 or older (11). Despite this, only 15% of heart failure studies have included patients over 80 years of age (12).

There is evidence from surveys of medical and related staff that the absence of relevant outcome data makes prescribing and treating older people difficult and impacts on equity of care (13). This is regarded as a problem across much of Europe (14). In a study of more than 500 clinicians in nine European countries, 87% of respondents concluded that exclusion on the grounds of age alone was unjustified and 79% felt this caused difficulties for prescribers. Respondents from most countries felt that changes to regulation were needed to reform current practice (14).

2 The evidence for under-representation

Evidence from published studies has shown that trial populations have not matched the age range of the clinical population (<http://www.predicteu.org/Reports/PREDICT_WP1_Report.pdf>). In a meeting of an expert committee of the European Medicines Agency, it was stated that in this context, ‘The drugs we are using in older people have not been properly evaluated’ (15). A consequence of age not matching the actual treatment group is that gender also will not be appropriate. Women survive into old age more commonly than men and so failure to recruit in these age groups will also disproportionately impact women. The PREDICT group investigated this by examining currently ongoing studies in the area of heart failure therapy as a key disease process for older people (16). Access to the European Medicine Database of ongoing studies was very difficult at the time; therefore data from ongoing clinical trials of heart failure were extracted from the World Health Organization Clinical Trials Registry Platform (<http://www.who.int/ictrp/publications/en/>). The key outcome measure was the proportion of trials excluding patients by an arbitrary upper age limit or by other exclusion criteria that indirectly limited recruitment of older individuals. Exclusion criteria were categorized as justified or poorly justified. Of 251 trials investigating treatments for heart failure, 64 (25.5%) excluded patients by an arbitrary upper age limit not related to other factors. Exclusion of older people was significantly (p = 0.007) more common in trials conducted in the European Union (31 of 96, 32.3%) than in the United States (17 of 105, 16.2%). Overall, 109 trials (43.4%) on heart failure had one or more poorly justified exclusion criteria that could limit the inclusion of older individuals due to the association of these factors with age.

Exclusion of older people in ongoing trials was independently examined in 206 studies of treatment for Parkinson’s disease. It was concluded that exclusion on the basis of upper age limit was common, especially in smaller studies, with 49% of studies having a mean upper age limit of 79 years (17).

3 Older people and clinical trials—barriers and promoters

3.1 Barriers

There are many reasons why older people may be excluded from clinical trials (Box 16.1). Exclusion of the elderly may be solely based on arbitrary age limits (18,19). Sometimes this is to satisfy a perceived or real requirement of ethics committees. This appears to be more common in interventional studies and, while possibly decreasing, particularly in the USA, it continues to be a major problem in Europe (20). There is evidence that ethics committees may be inconsistent. This is especially so when the area they are considering involves complex regulatory regimes such as the Mental Capacity Act, which deals with adults in England and Wales who lack capacity to give informed consent. Complexity may lead to exclusion of people who lack capacity, as investigators quickly learn which areas of research lead to confusion and delays in what is already a lengthy permission process (21).

Box 16.1 Barriers to research in older people

arbitrary age limits

inappropriate exclusion criteria that disproportionately affect older people

lack of resources to cope with cognitive deficits

restrictions on co-morbid conditions

restrictions on polypharmacy

complex regulation, not designed with research in mind

logistical issues, such as loss of income, timing, and transport

lack of provision for carers

under-resourcing to provide for the extra time needed to recruit and gather data from older people

lack of appreciation of extra needs by funding bodies

physician resistance to study protocol

communication barriers between community and hospital care

lack of recognition for clinicians involved.

In the UK it is often regarded as acceptable for a close family member to make a decision for the patient and proceed on that basis. This is helpful, but does mean that a suitable person may not be available and/or needs to be found in a timescale that fits in with the project’s requirements. An independent mental capacity advocate (IMCA) might be appointed for patients who lack capacity and do not have someone who can act as a consultee. Unfortunately, the IMCA role is not designed with research facilitation as a primary aim. A guidance document on such appointments was issued in 2008 (22). The IMCA could be a doctor independent of the research team. Although this may be the preferred route in emergency situations, it is not always easy to show that a nominated consultee is totally independent. The IMCA does not affect research that is a clinical trial as laid out in the Medicines for Human Use Regulations 2004, which include similar principles for the involvement of a ‘legal representative’.

Canadian studies of oncology patients involved in clinical trials found that age was an important barrier to recruitment, even though some older people may be easier to recruit because they lack the time pressures of those who are employed and are more prepared to attend follow-up visits. Physicians’ perceptions, protocol eligibility criteria with restrictions on co-morbid conditions, and functional status to optimize treatment tolerability are the most important reasons resulting in the exclusion of older patients. Other barriers include the lack of social support and the need for extra time and resources to enrol these patients (23).

In the UK, the research approval process is a major hurdle, and many believe that the simplest route to data is the best. Co-morbidity and polypharmacy complicate studies, increasing the numbers needed to recruit to an adequately powered study. The need to cope with transport and to facilitate communication in patients who may have impaired hearing or sight may increase the cost, time, and the skills needed to conduct a study (24). There is also a need to protect vulnerable subjects from adverse events in a study of a treatment that has unpredicted consequences. Therefore, some would argue that a simple initial study followed by a later study covering the needs of the elderly is the safest compromise. Unfortunately this depends on a further study actually being completed, which is difficult to guarantee without some form of legal compulsion.

In his systematic review for PREDICT (3), Andrew Beswick identified 19 studies reporting promoters and barriers to participation of older people in clinical trials extending back 20 years. These included the lack of an obligation for pharmaceutical companies to promote randomized controlled trials in older people and even lack of personal recognition for the physician involved.

3.1.1 Physician endorsement

Physicians’ perceptions of the implications of the trial for the patient are important (25,26). Physicians may not wish their patients to participate because the trial process requires extra time and resources or related practical difficulties. Alternatively, trials may have implications for loss of professional autonomy, impact on the patient-physician relationship, or present consent difficulties (24,27). In three surveys of patients with cancer, the endorsement of an oncologist was reported to be a factor in trial participation (28). Physicians might advise against participation if they felt their patients would show poor compliance (29) or if they thought a therapy was not the right one.

tIn some systems, insurance issues—either with regard to the study or health care—may present difficulties (27,30). Such practical barriers include insurance companies’ needing to agree, on an case-by-case basis, to the protocol for a study taking place in a private hospital. The idea is that the companies would want to weigh the risks of any extra provision they may have to cover as a result of the trial. In the USA, there was a change in 2000 in the Department of Health and Human Services medicare reimbursement policy, which allowed payment of routine care costs of older patients in clinical trials.

3.1.2 Transport and logistical barriers

Obstacles may be presented by financial or employment implications, such as transport costs, loss of work time, and inconvenience for a carer. Sometimes the patients themselves play a carer’s role, which could be compromised by a study.