This chapter describes the recruitment and follow-up of study participants, by site, and provides a baseline description of participants’ characteristics as well as baseline and follow-up values for the following outcome measures: the NPI, the EQ-5D visual analogue scale, the BADLS, the GHQ-28 (total and domain scores), the MMSE and the DEMQOL scale (total and selected domain scores). We present the data primarily to characterise the sample and document the extent of follow-up. Finally, serious adverse events occurring during the study are described.
Recruitment
The recruitment target was 44 dyads (person with dementia plus a carer), 11 from each practice. The number of dyads recruited was 29; 14 were recruited from the two north-east practices, nine from the Norfolk practice and six from the London practice. Recruitment was halted at the Norfolk practice to deal with a backlog of case management work and this practice would probably have achieved its target had more time been available to the nurse case manager. Case identification using the Quality and Outcomes Framework (QOF) dementia register was supplemented by searches of electronic medical records to identify those taking cholinesterase inhibitors who were not on the QOF dementia register. Additional searches for patients with symptoms suggesting possible dementia (memory loss, confusion) allowed medical records to be checked for evidence that a formal diagnosis had been made but had not been added to the patient record. shows the derivation of the participant sample for the feasibility study.
Recruitment flow diagram.
Of those patients not living in care homes, 45 [33%, 95% confidence interval (CI) 29% to 48%] met all of the criteria for inclusion in the study apart from providing informed consent. In total, 28 of these (62%, 95% CI 46% to 76%) agreed to participate.
Recruitment and follow-up by site
In London, six participant dyads (12 participants – six people with dementia and their main carer) were recruited between 17 August 2012 and 18 December 2012. Three potential participants agreed to speak with the researcher but declined to participate in the study. Reasons provided for this were not having the time to participate (two people) and not wanting their relative with dementia exposed to research questioning (one person). In London, a female patient was referred to the study and agreed to meet one of the researchers but it became apparent that she was unlikely to have dementia. The researcher checked with the surgery, which had placed her on the QOF dementia register without a diagnosis of dementia confirmed by secondary assessment. The researcher wrote to her to thank her for her time and withdrew her from the study. Follow-up appointments for the recruited participants in London were completed at 5 months. Six dyads were successfully contacted and follow-up data were collected.
In Norfolk, nine participant dyads (18 participants – nine people with dementia and their main carer) were recruited between 7 August 2012 and 24 September 2012. Three patients who were invited to participate returned the opt-in form stating that they would not be interested in taking part but gave no details about why they were declining participation. In Norfolk, all 18 participants recruited into the study completed a follow-up appointment at 5 months.
Thirteen dyads plus one carer-only participant were recruited in the north-east. Ten dyads were recruited through the larger practice. Recruitment took place between 30 July 2012 and 27 November 2012. Only one patient returned the opt-in form stating that she was not interested in participation.
A further four participants were agreeable to their details being passed to the researcher; however, either the patient or the nominated carer declined participation when approached by the researcher. Reasons for declining participation included patient illness/hospital admission and other family or care commitments. Despite multiple attempts, the researcher was unable to contact an additional two patients who had expressed an interest in participation through the GP contact. shows the derivation of the denominator for outcome assessment at all sites.
Recruitment of dyads and follow-up by site
Baseline data collection
Baseline data were collected from all dyads (nine from Norfolk, six from London and 13 dyads and one carer from the north-east). All nine dyads from Norfolk and all six from London were followed up at 5 months, with complete data collection. Ten of the 13 dyads from the north-east were successfully followed up at 5 months (outcome data were collected from four and the other six indicated that they were willing to provide it). Two dyads in the north-east indicated that they were not willing to provide outcome data (one patient had moved into residential care and her carer felt that the patient was not capable of participating; one patient did not receive the intervention and the carer felt that data collection would be too upsetting for the patient) and one further patient was lost to follow-up.
Demographics of participants at baseline
and show the characteristics at baseline of the carers and the people with dementia, respectively.
Characteristics of carers (n = 29)
Characteristics of patients with dementia (n = 28)
Outcome measures
For each variable we provide an indication of the distribution of responses at baseline, usually in the form of a box and whiskers plot. Additionally, we present descriptive statistics for each of the measures of outcome and quality of life used in the pilot study. The purpose behind this is to facilitate the planning of any future studies. In particular, measures of variability are likely to inform sample size calculations; it is not intended that the data be used to make comparisons between the particular sites that participated in this study.
Neuropsychiatric Inventory
The NPI was chosen as the likely primary outcome measure for a definitive trial. The NPI assesses 10 behavioural disturbances occurring in dementia patients: delusions, hallucinations, dysphoria, anxiety, agitation/aggression, euphoria, disinhibition, irritability or emotional lability, apathy and aberrant motor activity. Higher scores suggest higher levels of disturbance. The distribution of total scores for the patients recruited to the feasibility study is provided in and .
Neuropsychiatric Inventory total scores: descriptive statistics
Neuropsychiatric Inventory total scores: box and whiskers plot of baseline values.
In (and subsequent box and whiskers plots), the horizontal line within each box corresponds to the median value, the lower and upper edges of the box correspond to the 25th and 75th percentiles and the whiskers indicate the range, except for any outlying values, which are indicated by circles and or stars. plots the NPI scores at 5 months against the baseline scores.
Plot of NPI scores at 5 months against scores at baseline.
Most of the scores at 5 months are greater than the scores at baseline. There are four exceptions, three of which correspond to responses from London carers or patients. The mean increase in NPI total score is 4.3 (95% CI –0.31 to 8.56) (based on 50,000 bootstrap samples).
European Quality of Life-5 Dimensions visual analogue scale
This general health rating scale (0–100) was completed by carers. Higher values suggest a higher quality of life. shows the descriptive statistics for the EQ-5D visual analogue scale at baseline and at the 5-month follow-up and shows a box and whiskers plot of baseline values.
European Quality of Life-5 Dimensions visual analogue scale: descriptive statistics
European Quality of Life-5 Dimensions visual analogue scale: box and whiskers plot of baseline values.
Bristol Activities of Daily Living Scale
The BADLS was designed specifically for use in patients with dementia and assesses 20 daily living abilities. A score of 0 suggests total independence and a score of 60 suggests total dependence. shows the BADLS scores at baseline and at 5 months’ follow-up and shows baseline values as a box and whiskers plot.
Bristol Activities of Daily Living Scale total scores: descriptive statistics
Bristol Activities of Daily Living Scale total scores: box and whiskers plot of baseline values.
General Health Questionnaire
The GHQ-28 has possible scores across four domains in the range from 0 to 84, with higher scores suggesting higher levels of psychological disorder. The four domains are (A) somatic symptoms, (B) anxiety and insomnia, (C) social dysfunction and (D) severe depression. shows the GHQ-28 total scores at baseline and at 5 months’ follow-up and shows the GHQ-28 total scores at baseline as a box and whiskers plot. shows the domain scores at both time points and shows the baseline domain scores as a box and whiskers plot.
General Health Questionnaire total scores: descriptive statistics
General Health Questionnaire total scores: box and whiskers plot of baseline values.
General Health Questionnaire domain scores: descriptive statistics
General Health Questionnaire domain scores: box and whiskers plot. A, somatic symptoms; B, anxiety and insomnia; C, social dysfunction; D, severe depression; and Tot, total.
Mini Mental State Examination
The range of scores on the MMSE is from 0 to 30, with lower scores suggesting worse cognitive impairment. shows the MMSE scores at baseline and at 5 months’ follow-up and shows the baseline MMSE scores as a box and whiskers plot. A MMSE score of < 10 suggests advanced dementia ().
Mini Mental State Examination scores: descriptive statistics
Mini Mental State Examination scores: box and whiskers plot.
Mini Mental State Examination category at baseline
Dementia Quality of Life scale
The DEMQOL is a quality of life measure in which higher scores imply better quality of life. Participant total scores at baseline and at 5 months’ follow-up are shown in ; displays baseline values as a box and whiskers plot. The DEMQOL scale includes five domains, three of which are reported in (‘feelings’ = health and well-being; ‘memory’ = cognitive function; and ‘everyday’ = daily activities and looking after yourself). Baseline domain scores are provided in as a box and whiskers plot.
Dementia Quality of Life scale total scores: descriptive statistics
Dementia Quality of Life scale total scores: box and whiskers plot.
Dementia Quality of Life scale domain scores: descriptive statistics
Dementia Quality of Life scale domain scores: box and whiskers plot.
Safety: serious adverse events
In London, two serious adverse events were reported. One involved a carer being admitted to hospital and one involved a patient (from a separate dyad) being admitted to hospital. These admissions were the result of falls and were unrelated to the study but were expected because of the nature of the patient and carer population. In the north-east, four serious adverse events were reported. One patient was hospitalised twice, both times because of a fall. Another patient suffered a fall and one patient was hospitalised with suspected pneumonia. Again, these serious adverse events were expected because of the age of the patients recruited. No serious adverse events were reported in Norfolk.
Costs
Case management salary costs for WP2 amounted to £52,890 across all sites. Training and mentoring by an Admiral nurse cost £5201 and £6273, respectively.
