In order to guarantee that the future care provided to our ageing populations is grounded in the best quality evidence, it is imperative that researchers ensure their study populations are representative and include the oldest old, the most rapidly expanding sector of the population. Older people, especially those over 75 and those in care homes, are often unjustifiably excluded from studies [24
]; this must be addressed by both the funding commissioners and the research community by tackling cultural/investigator bias and unjustified age limits. Our recruitment rates were good considering the age of the sample and the extensive assessment involved. These rates also compare favourably with other UK longitudinal studies of ageing, which have often involved younger cohorts [3
], and with similar studies internationally [28
]. We accept that variation in local circumstances may be an important factor and that the procedures used here were suitable for the particular population under study and may need suitable modification for other populations. Although this is a limitation of the study, we feel that researchers in other areas will readily appreciate the need to modify procedures where necessary.
Our experiences with residents of care homes confirmed previous recommendations [12
] that the additional time required to recruit this very important sub group of older people is substantial. As Zermansky et al. (2007)[14
] commented, "conducting research with care home residents is beset with constraints and complexities that can impair progress.....an extraordinary amount of time and resources are needed to overcome them". It is therefore essential that such resources be costed into the project funding and that, in order to inform these costings, procedures for recruitment from care homes be tested in a pilot study.
Throughout this paper, we have outlined issues we encountered in involving and retaining the oldest old in a large cohort study of health and ageing in one area of the UK and put forward approaches we found useful to successfully address these. In summary, important factors to consider when engaging the oldest old in research include:
• Initial participant approach. An initial invitation from a known health care professional may lead to more confusion than a letter from the study team. Photographs of researchers are useful additions to study information sheets. Adequate time for recruitment and data collection should be factored into planning.
• Engagement with family and care home staff. Additional time and resources are needed, especially with the oldest old, to engage with family carers and other gatekeepers such as care home staff for those people living in care. Researchers should be trained and supported to ensure that complex situations are handled with sensitivity.
• Significant cognitive impairment is an issue; due consideration must been given to the relevant ethical and legal issues, including assessment of capacity and consent procedures. The use of protocols to assess capacity and to secure consent should be developed. Training for researchers in the assessment of capacity and seeking informed consent should be mandatory.
• Minimising participant burden. Home visits are preferred by this population and may enable the participation of those who are frail or housebound. Consequently researchers' travel costs, specialised portable equipment and availability of chaperones may need to be costed into the research. The need for, and the cost of, several, shorter visits to assess capacity, secure consent and collect all relevant data should be anticipated.
• Participant withdrawal. Researchers should endeavour to make the research process as flexible as possible and may wish to consider the option of study suspension as an alternative to withdrawal as often the participant does not wish to withdraw permanently but cannot accommodate study visits at that particular time due to other commitments.
• Communication with participants after study cessation. The oldest old may derive benefits, such as opportunities for increased socialisation, from participation in research. The process of study close-out should be considered at the onset so that participants are not left with a feeling of loss. A communication strategy to update participants about the study progress and findings after their involvement ceases can be helpful.
• Engagement with stakeholders and the media. Researchers should consider the need to engage with not only local clinicians, but also with older people's voluntary organisations, the media and the local press. This should be proactively built into the study recruitment protocol.