Recruitment
The cohort will be recruited from general practices in Newcastle and North Tyneside Primary Care Trusts. Eligible individuals will be all those who turn 85 during the year 2006 (i.e. born in 1921) and who are registered with any Newcastle or North Tyneside Primary Care Trust general practice. All individuals who meet these inclusion criteria will be invited to participate, whether living at home or in an institution, and regardless of their state of health (except for exclusion by their general practitioner (GP) due to end stage terminal illness). Potential participants will be identified through the primary care trusts' GP patient lists and we will liaise with GPs to cross-check these lists (to exclude recent deaths and the terminally ill) and to request a 'letter of support'.
The study team will send each potential participant a letter of invitation together with an information pack, photographs of those research staff involved in home visits and a 'letter of support' from their GP. The letter of invitation will be followed after one week by a telephone call or home visit from a research nurse to discuss the study in more detail. Previous experience in similar studies involving very old people has shown that direct telephone contact or visiting are the only reliable ways of finding out whether individuals are interested in participating in research. As many very old people are frail, they may find posting reply slips or making the first telephone call themselves too much of a burden. No pressure is put on the person to participate; this procedure has been adopted purely as a way of ensuring the older person receives all the relevant information to enable them to make an informed decision about participation. Once an individual has shown an interest in participating, an appointment is made to visit them in their own home. The purpose of this visit is to ensure that they have received and understand the study information and the implications of their involvement. If they wish to participate, written informed consent is then obtained.
As around 15% [
34] of the proposed sample may have significant cognitive impairment, due consideration has been given to the ethical issues this raises. It is important not to exclude such people from scientific research as, so doing, could prevent this group from benefiting from advances in medical understanding. However there is a need to carefully consider the consent process. The Mental Capacity Act 2005 provides a statutory framework 'to empower and protect vulnerable people who may not be able to make their own decisions' [
38]. At the forefront of this framework is the principle that capacity of an individual to consent must be assumed unless it is proved otherwise. During the first visit to a potential participant, the research nurse will establish their level of capacity utilising a consent checklist and consent pathway, based upon information from the individual, family and significant others. Each participant will be asked to give consent appropriate to their level of understanding ranging from written full informed consent to verbal or non-verbal communication, of which account will be taken in determining willingness to participate. In those individuals found to be without capacity to give full informed consent, proxy assent/consultee approval will also be sought from a close relative or, in those potential participants without relatives, from an immediate carer. They will be asked to use their knowledge of the participant, in the past and the present, and will be required to state that, in their opinion, the participant would have had no objection to entering the study when cognitively intact, and would not be caused undue distress by participation. The consent protocol incorporates the notion that consent is a process and as such requires ongoing confirmation prior to and within each contact. If a participant is found to have lost capacity during the course of the study, consent is not deemed as being enduring and a process of gaining proxy assent/consultee approval is initiated, whilst suspending all contacts.
Assessment
Participants will be visited in their current residence (own home or institution) by a research nurse. If a potential participant is temporarily in hospital at the time of recruitment (1% of those eligible to participate in the pilot study), assessment will be deferred until discharge. Home based assessment is vital for high recruitment levels in this age group; in our pilot study over half of the sample would have been unwilling to attend hospital for assessment.
The baseline assessment protocol entails a detailed multi-dimensional health assessment by means of:
a). Questionnaires: socio-economic status; family data; physical health (global health status, longstanding illness, angina, shortness of breath, falls, generalised pain, joint pain, fractures, incontinence, vision and hearing); psychological health (depression); disability; diet; oral health; lifestyle (smoking, alcohol and exercise); social support and participation; use of health and social care.
b). Measurements and function tests: weight, bio-impedance (body composition-fat and water), demi-span, waist and hip circumference, tooth count, blood pressure, hand-grip strength, walking test (timed 'up and go' test), cognitive function (mini-mental state examination and computerised assessment of memory and attention (CDR battery)), electrocardiogram, spirometry and oximetry.
c). Blood tests: blood samples, taken after an overnight fast where possible, will be analysed for:
1. Routine haematology and biochemistry: full blood count; creatinine and electrolytes; liver panel; bone panel; glucose; glycosylated haemoglobin.
2. Lipid profile: cholesterol, triglycerides, high and low density lipoproteins, apolipoproteins (A1 and B).
3. Thyroid function: free T4, free T3, reverse T3, TSH and TPO antibodies.
4. Inflammatory markers: High sensitivity CRP, rheumatoid factor, cytokines (TNF-alpha and Interleukin 6).
5. Cortisol
6. Nutritional markers: Vitamins B2, B6, B12, C and D, ferritin, red cell folate and homocysteine.
7. Biomarkers: DNA repair capacity, telomere length, F2-isoprostane (marker of oxidative stress).
8. Markers of immunosenescence: T cell oligoclonality and lymphocyte subpopulation distributions (senescent T-cells, memory T-cells and NKcells).
In addition, DNA, RNA and plasma will be stored for future analyses (including genetic, proteomic, metabolomic and allostatic load markers concerned with ageing [
39]).
Information will be gathered during three nurse visits each lasting about one hour and 30 minutes and one additional nurse visit to collect the fasting blood samples; visits to individual participants will take place over one month. Data will be entered directly onto a tablet style laptop computer. For those participants who are particularly frail, there will be the option to spread the assessments over more visits, to have a shorter assessment in total and for much of the interview to be conducted with an 'informant' i.e. a relative/carer who knows the participant well. With participants' consent, their GP will be informed about their participation and of any abnormal assessment findings of immediate clinical relevance. In addition participants' GP records will be reviewed for information on medication, key diagnoses and use of GP services. For those individuals who decline to participate in the full assessment, there will be the option to consent to review of records only.
The full protocol was tested during the pilot study and shown to be fully practical and practicable. The schedule was shown to be acceptable to participants; 96% rated their involvement in the study as positive or very positive and 93% of those recruited completed all the nurse visits. Safety issues have been addressed; research nurses will have enhanced Criminal Records Bureau clearance and carry photo-identification on visits, participants will be sent photographs of the research team with their invitation letter and local police will be informed about the study. For staff safety, the location of staff involved in community visits will be monitored using an automated telephone call-back system.
It will not be possible to check inter-rater reliability for home assessment data collection due to the participant interview burden this would entail. However reliability of the data extraction from GP records will be examined in a core set of variables which cover the range of difficulty and modes of extraction from GP records. From the pilot data, where 3 individuals extracted data independently from the same GP records, we found the intraclass correlation coefficient (ICC) to vary from 0.55 for number of heart attacks to 0.99 for number of consultations. We estimate therefore that 9 nurses extracting information from the same set of 32 notes will have 80% power with a type I error rate of 5% to show at least moderate agreement (ICC>0.4) for the majority of variables extracted.