Background
Dementia is a major public health problem with enormous costs to society [
1]. It is an incurable progressive disease with devastating consequences for both patients and their relatives. The estimated prevalence rate of dementia among older adults aged 65 to 95 is 6.6% [
2]. Over the next years the number of demented older persons will increase substantially as a result of aging populations [
3].
Initially, informal caregivers, such as relatives, neighbors and friends, care for most patients with dementia. Caregiving is generally unplanned and most informal caregivers gradually adopt their role because of the insidious nature of dementia [
4]. However, informal caregivers often experience adverse psychological, physical, social, and financial consequences [
5]. Compared with non-caregivers, they live shorter and report more depressive symptoms [
6,
7]. Besides, caring for a demented person is marked by losses of previous roles in a relationship. Moreover, many caregivers reduce or give-up the time spent on paid jobs and social activities [
4].
Timely detection of dementia is important for both patients and their caregivers as it enables care support and prepares future care [
8]. However, there is evidence of underdetection [
9,
10] and diagnostic delay [
11]. An important patient related barrier to timely recognition is the absence of a request for help. This absence can be attributed to denial, labeling cognitive impairment as an accepted aspect of normal ageing, lack of awareness of the disease process, or the idea that nothing can be done [
12,
13]. In contrast to conventional care, pro-active care with timely detection followed by structured care focusing on both demented patients and informal caregivers, may be more suitable for this vulnerable group. So far, randomized controlled trials of such pro-active disease management systems have not been reported. Yet, up till now, to assist informal caregivers of demented older adults, several psychosocial support programs have been developed, such as support groups, respite care, stress-management, social skills training, psycho-educational groups, and case-management. On the whole, multicomponent interventions that provide caregivers with diverse services and supports, and individually tailored interventions showed larger effects on caregivers' well-being than other, narrowly focused interventions [
14‐
18]. We use the concept 'sense of competence' to denote the caregiver's feeling of being capable to care for the demented person.
Interventions showed increased caregivers' sense of competence [
19,
20], stabilized caregivers' well-being [
21,
22] to sustained benefit in reducing depressive symptoms [
23], changed caregiver's appraisals of patient behavioral problems [
24], and, lastly, postponement of patients' institutionalization [
19,
20,
25‐
28], although there is lack of strong findings in general [
18,
29,
30]. Trials on case-management, showed a deferral or no reduction in patients' institutionalization rate [
27,
31], and on the whole did not impact caregivers' levels of depression and burden, in spite of small reductions at some sites [
21]. Few studies have performed economic evaluations of interventions for community-dwelling dementia patients [
32‐
34]. Cost-effectiveness analyses and cost-utility analyses are even rare [
35].
An innovative initiative to support dementia patients and their caregivers was set up by the Department of General Practice of the VU University medical center, GPs and district nurses in West-Friesland, the Netherlands. We developed a pro-active program, in which the key elements consist of timely detection of dementia symptoms followed by case-management by district nurses among detected patients and their primary informal caregivers. A randomized controlled trial (RCT) is performed to observe effectiveness and cost-effectiveness of case-management.
The objective of this paper was to describe the study protocol of this RCT among informal caregivers of men and women aged 65 years or over with dementia symptoms who live at home, and the men and women they take care of. The main research questions of this RCT concern whether case-management is more effective than usual care in improving caregiver's sense of competence, and whether case-management is cost-effective compared to usual care when assessed from a societal perspective. A secondary research question is whether case-management is more effective than usual care in improving caregiver's quality of life, caregiver's psychological well-being, caregiver's burden, patient's quality of life, and in decreasing hospital days, days until institutionalization and death of the patients.
Discussion
In this paper we described the study protocol of an innovative RCT that evaluates case-management by district nurses to primary informal caregivers of men and woman aged 65 or over with dementia symptoms who live at home, and the older men and women who receive informal care. This is one of the first trials on case-management that includes an economic evaluation. Moreover, it concerns a tailor-made intervention in early-detected patients with dementia symptoms and their caregivers. In addition, the detection method of patients with dementia symptoms preceding recruitment of these patients and their informal caregivers is unique. A large general practice population of older patients was approached by mailed questionnaires. Particular strengths of our study protocol are the randomization approach, in which allocation concealment involves an external independent person, and methods used to enhance the quality of measurements such as assessors who are blinded to group assignment and training of assessors. Another strength is the possibility to visualize delivery of the intervention and usual care by cost diaries. Cost diaries might also provide insight in factors related to the intervention process that may influence the effectiveness of case-management.
Below, we describe design features that address potential threats to reliability and validity. Firstly, selection of participants may limit generalization of the results of this study as selective non-response of older adults, selective refusal of caregivers, and selective dropout are possible. Non-responding older adults in other studies have been observed to have higher rates of functional and cognitive impairment [
61,
62]. To limit this potential selection bias we will send personalized invitation letters by GPs and provide reminders to initial non-responders. This strategy has shown to be effective [
63]. Furthermore, we anticipate that caregivers check mail of cognitively impaired individuals and provide help with filling out as inhabitants are informed about the project by a newspaper article. Selective refusal of caregivers to participate might be assumed as some caregivers will label cognitive impairment as an accepted aspect of normal ageing, or do not experience adverse consequences of caregiving. Possibly, such caregivers will refuse more often than other caregivers. The same might be assumed about severely burdened caregivers who could be afraid to become even more burdened with participating in the project's measurements. To limit such selective refusal, interviewers will contact potential participating caregivers after screening to inform them about the project before sending personalized invitation letters to them. To prevent selective drop-out of severely burdened caregivers and severely disabled patients, appointments for measurements are made by one fixed interviewer on times and locations suitable for the participants.
Secondly, two situations may cause information bias. Firstly, bias may occur as cognitively impaired subjects without insight may fill out questionnaires. However, as we assume that detected patients suffer mainly from mild or moderate dementia symptoms, and insight is mainly preserved in these subjects, this bias probably will be limited. Secondly, in the economic evaluation, caregivers are allowed to provide assistance or to fill out the EQ-5D when patients are unable to fill out this questionnaire. This may lead to information bias, as it is known that agreement on the EQ-5D between patients and caregivers is poor [
64]. However, this bias probably will be limited as well, as we assume that detected patients suffer mainly from mild or moderate dementia symptoms, and most patients will fill out the questionnaire themselves.
Thirdly, contamination could bias results of this study as we choose to perform randomization on patient level. However, influence of contamination on results is unlikely as participants of the usual care group have no access to particular activities of the intervention (e.g. family meetings, RAI-HC assessment, guide for informal caregivers). Nevertheless, it is possible that participating GPs are encouraged by the project to give more attention to patients with dementia symptoms and their informal caregivers participating in the usual care group.
Lastly, we expect heterogeneity in study subjects because response to interventions may be different depending on caregiver circumstances. In combination with the relatively small sample size of approximately 100 participants, this heterogeneity may make it hard to interpret the outcomes. However, increasing the sample size is not feasible. Therefore, we will visualize distribution of characterizes over comparison groups to estimate the influence of this heterogeneity on outcome measures. Moreover, cost diaries will detect heterogeneity in received care within the usual care group as well as in the case-management group.
The results of this RCT will provide valuable information for health professionals and policy makers on effectiveness and cost-effectiveness of timely tailor-made case-management for patients and their informal caregivers. Moreover, positive effects will challenge current health care systems to move to more pro-active approaches for this group. In case of proven effectiveness and cost-effectiveness, we recommend implementing this case-management intervention into usual healthcare. The results of this study will be available in autumn 2006.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
All authors have read and approved the final version of the manuscript.
AJ is the principal investigator and writer of this manuscript
HH wrote the study protocol and supervises the planning and project
HM wrote the study protocol and supervises the planning and project
GN supervises the planning and project
MB supervises the economic evaluation
JB performs the economic evaluation
AP trained nurses and assisted in putting together the care-program
WS supervises the project