Guidelines for psychosocial interventions in dementia care: a European survey and comparison
The effectiveness of psychosocial interventions in treating people with dementia and their carers is increasingly emphasised in the literature. Dementia guidelines should summarise the scientific evidence and best practice that is currently available, therefore, it should include recommendations for psychosocial interventions. The aims of our study were (1) to collate dementia guidelines from countries across Europe and to check whether they included sections about psychosocial interventions, and (2) to compare the methodological quality and the recommendations for specific psychosocial interventions in these guidelines.
The European dementia guidelines were inventoried. The methodological quality of the guideline sections for psychosocial interventions was assessed with the (AGREE) Appraisal of Guidelines Research and Evaluation instrument. The recommendations for specific psychosocial interventions were extracted from each of these guidelines and compared.
Guidelines for psychosocial interventions were found in five of 12 countries. Guideline developers, methodological quality and appreciation of available evidence influenced the inclusion of psychosocial interventions in dementia guidelines from Germany, Italy, the Netherlands, Spain and the UK. The UK NICE SCIE guideline had the best methodological quality and included the most recommendations for psychosocial interventions. Physical activity and carer interventions were recommended the most across all guidelines.
The inclusion of psychosocial interventions in dementia guidelines is limited across Europe. High-quality guidelines that include psychosocial interventions and are kept up to date with the emerging evidence are needed. Throughout Europe, special attention to the implementation of evidence-based psychosocial care is needed in the next few years. Copyright © 2011 John Wiley & Sons, Ltd.
The value of psychosocial interventions in treating people with dementia and their carers is increasingly emphasised in the literature. The goals of psychosocial interventions are to improve the quality of life and to maximise function in the context of existing deficits (APA, 2007). Scientific evidence for the effectiveness of this type of intervention in dementia care is growing because such interventions have shown a positive impact on cognition, quality of life and family carers as well as suppressing neuropsychiatric symptoms and associated distress (Spector et al., 2003; Teri et al., 2005; Graff et al., 2006; Mittelman et al., 2006). Psychosocial interventions are not limited to specific types of dementia, and no serious side effects have been reported.
Pharmacological treatments have shown some efficacy, but contrary to psychosocial interventions, antidementia drugs are not effective for all types of dementia and are not tolerated by all patients (Wild et al., 2003; Birks, 2006; McShane et al., 2006). Furthermore, the efficacy and safety of drug therapies for neuropsychiatric symptoms have been questioned over the last decade (Ballard et al., 2009a; Ballard et al., 2009b; Gauthier et al., 2010). Psychosocial interventions are recommended as a safe first-line treatment for neuropsychiatric symptoms and are efficacious at all stages of dementia (Aarsland et al., 2005; Kverno et al., 2009). Psychosocial interventions for carers are effective in postponing and decreasing the odds of institutionalisation, but similar results for pharmacological treatments are lacking (Brodaty et al., 2003; Mittelman et al., 2006; Spijker et al., 2008).
The use of scientific evidence in healthcare decision-making is not automatic, and gaps between evidence and decision-making are present at all healthcare levels (Straus et al., 2009). Guidelines are an important tool for implementing evidence in daily practice. Ideally, they should summarise the scientific evidence and best practice that is currently available (Grol et al., 2005; Brouwers et al., 2009). Guidelines for diagnosing and treating dementia have been developed and published in several European countries. Besides scientific evidence, factors such as the healthcare system and professional culture may influence the inclusion of psychosocial interventions in these guidelines.
The organisation and financing of dementia care (particularly social support) is often fragmented and differs between and within European countries (Alzheimer Europe, 2006; Alzheimer Europe, 2007). Some countries organise dementia care nationally, but others depend on regional and local policies and financing (Alzheimer Europe, 2006; Alzheimer Europe, 2007). In the latter case, national dementia guidelines, if they exist at all, may include only general and nonspecific recommendations for psychosocial interventions to fit all regions. Or development of dementia guidelines may rely on local initiatives. Furthermore, medical specialists diagnose most cases of dementia, but they tend to focus their treatments on pharmacological interventions (Rimmer et al., 2005). Clinical guidelines developed by medical specialists might not include recommendations for psychosocial interventions.
All Europeans with dementia and their carers should have access to safe care, that optimises their quality of life and increases their chances of staying at home as long as possible. Not including psychosocial interventions in dementia guidelines could have consequences for the quality of dementia care.
Recently published European dementia guidelines should therefore include evidence and recommendations for psychosocial interventions that parallel the emerging evidence base. The aims of our study were (1) to collate dementia guidelines from European countries and check whether they include sections about psychosocial interventions, and (2) to compare the methodological quality and recommendations for specific psychosocial interventions of these guidelines.
We used a pan-European, multiprofessional clinical research network, INTERDEM (Timely Detection and Intervention in Dementia; http://interdem.alzheimer-europe.org/) to gather information about guidelines and consensus papers for psychosocial interventions in dementia in Europe between May and October 2006. INTERDEM members in the following countries were sent an email with a request to gather relevant national guidelines: UK, Spain, Netherlands, Belgium, France, Germany, Ireland, Italy, Portugal, Switzerland, Greece, Poland, Sweden, Austria, Denmark and Finland.
We extracted the title and the names of the authors and intended users from each guideline and consensus paper that we received. We assessed the methodological quality of the guidelines that included at least one recommendation for a specific psychosocial intervention and that were published from 2000 onwards. We also extracted the recommendations for specific psychosocial interventions from each guideline and compared them. The guideline with the highest Appraisal of Guidelines Research and Evaluation (AGREE) scores was taken as a reference point. We compared recommendations treating carers and neuropsychiatric symptoms in more detail because of emerging evidence of the effectiveness of psychosocial interventions in both matters.
To measure the methodological quality of a guideline, we used the AGREE instrument, which was internationally developed and tested for appraising clinical practice guidelines (The AGREE Collaboration, 2003). This instrument assesses the quality of reporting and the guideline development. It provides an appraisal of the predicted validity of a guideline, i.e. the likelihood of a guideline achieving its intended outcome. It comprises 23 items that are unequally divided over six domains:
- (1)Scope and purpose (3 items), which cover the overall aim of the guideline.
- (2)Stakeholder involvement (4 items), which covers the extent to which the guideline represents the views of its intended users.
- (3)Rigour of development (7 items), which covers the process used to gather and synthesise the evidence and the methods for formulating and updating the recommendations.
- (4)Clarity and presentation (4 items), which cover the language and format of the guideline.
- (5)Applicability (3 items), which covers the likely organisational barriers and the behavioural and cost implications of applying the guideline.
- (6)Editorial independence (2 items), which covers the independence of the recommendations and acknowledgement of possible conflicts of interest of the guideline development group.
Each item was rated on a four-point Likert scale ranging from 4 (strongly agree) to 1 (strongly disagree).
Guidelines for psychosocial interventions were rated independently by at least two clinicians, practitioners or researchers in dementia care who were native speakers or who could read the guideline in its original language. We calculated standardised scores for each guideline, and for each domain of AGREE. We added the item scores of all appraisers to calculate the standardised scores as: [(obtained score − minimum possible score)/(maximum possible score − minimum possible score)] × 100%. The maximum and minimum possible scores = the number of items × the number of appraisers, either × 4 (strongly agree) or × 1 (strongly disagree). The ratings were based solely on the sections about psychosocial interventions. The only other guideline sections considered during the rating were the more general parts about the development of the guideline. Therefore, the scores do not reflect the quality of a guideline as a whole.
Information about published guidelines and dementia care consensus papers was received from 12 of the 16 countries in the INTERDEM network and supplementary Alzheimer's Society contacts. This information included 31 dementia guidelines and consensus papers from 8 different countries that were published between 1998 and 2006.
There were no recommendations for using psychosocial interventions in four countries (Finland, Belgium, Denmark and Sweden), although two guideline groups in Sweden had work in progress, which they expected to complete in 2009. In France, Switzerland and Ireland, consensus papers that included psychosocial interventions were published, but we found no recommendations for using psychosocial interventions in the dementia guidelines. The range of guidelines and consensus papers for dementia care and their content are described elsewhere (Vasse et al., 2008).
Seven guidelines from five countries were included for the assessment of methodological quality. They comprised one Italian guideline for clinical specialists (Caltagirone et al., 2005), one Spanish guideline for health and social care practitioners (Spanish Multidisciplinary Group, 2000), one German guideline for formal carers in institutional care (Bartholomeyczik et al., 2007), two UK guidelines for practitioners and service commissioners (NICE SCIE, 2006) and healthcare professionals (SIGN, 2006), and two guidelines from the Netherlands for general practitioners (Wind et al., 2003) and formal carers in institutional care (NIVEL, 2004a,b).
Table 1 gives the AGREE scores for all seven guidelines. Table 2 shows a comparison of the psychosocial interventions included in the guidelines. Table 2 only shows the interventions that at least one of the seven guidelines recommended. Both tables present the guidelines in order of highest to lowest score for the AGREE domain for rigour of development, which covers the searching for and selecting evidence and formulating and updating the recommendations.
Table 1. Appraisal of sections about psychosocial interventions in European dementia guidelines. Appraisal of Guidelines Research and Evaluation domain scores
|Scope and purposes||89||65||89||83||57||39||78||71|
|Rigour of development||90||79||79||52||45||43||38||61|
|Clarity and presentation||88||56||67||63||42||67||71||65|
Table 2. Comparison of recommendations for psychosocial interventions in different European guidelines
|Cognitive behavioural therapy||Yesm||—||—||Yesn||—||—||—|
Quality of guidelines
Overall, the scope and purposes domain scored highest on average of all the AGREE domains (mean score 71%, range 39–89%, Table 1). Thus, in general, the objectives, the clinical questions covered and the patients to whom the guideline is meant to apply were specifically described in most guidelines. In contrast, the applicability domain scored lowest overall (mean score 32%, range 6–89%). This means that the organisational changes and cost implications of applying the guideline recommendations were ignored or barely described in most guidelines.
The rigour of development domain scored 61% (range 38–90%) on average. The items rated the highest in this domain were about using systematic methods for finding evidence and the criteria for selecting the evidence. The low score items were the two items outlining the guideline updating and the external review process before publication.
The NICE SCIE guideline from the UK scored best in all domains for its coverage of psychosocial interventions; it had the best overall methodological quality in this context. The UK guidelines scored the highest for rigour of development. They used systematic methods to find and select evidence, explicitly linked recommendations to the supporting evidence, mentioned an external review process, and outlined an updating process for the guidelines.
The German guideline scored as high as the SIGN guideline for rigour of development, but did not mention an updating process. It scored under 50% only for the applicability domain.
The two Dutch guidelines scored below the mean domain score for rigour of development. Both explicitly linked recommendations to supporting evidence, but the guideline for general practitioners did not mention an updating process. The Dutch guidelines scores for the applicability domain were lower than their scores for the other AGREE domains.
The Italian guideline clearly described the methods for finding and selecting evidence, and it linked recommendations the supporting evidence. However, it did not provide a procedure for updating the guideline. This guideline had the lowest overall domain score, namely, 6% for applicability.
The Spanish guideline had the lowest score of all guidelines for rigour of development. It did describe how evidence was searched for and methods for formulating recommendations, but did not describe clearly how evidence was selected or set forth a procedure for updating the guideline. Across domains, this guideline had the lowest score for applicability and editorial independence.
Comparison of recommended psychosocial interventions
The psychosocial interventions that were most often mentioned across guidelines were physical activity, carer interventions, multisensory stimulation/snoezelen and reminiscence (Table 2). All guidelines that mentioned the first two interventions also recommended them. While guidelines mentioned light therapy, life review involving negative memories, psychomotor therapy and simulated presence, the guidelines either did not recommend them or advised against them. Three guidelines advised not using light therapy (Caltagirone et al., 2005; NICE SCIE, 2006; SIGN, 2006). One guideline advised against life review involving negative memories (Bartholomeyczik et al., 2007).
The NICE SCIE guideline had the best methodological quality and included the most recommendations for psychosocial interventions. All psychosocial interventions recommended by the SIGN guideline were also recommended by the NICE SCIE guideline. The UK guidelines were the only guidelines that recommended cognitive stimulation. The NICE SCIE guideline, however, included reality orientation and cognitive stimulation as one and the same intervention, whereas the SIGN guideline recommended them as two separate interventions.
Five of the six psychosocial interventions recommended by the Italian and German guidelines were also recommended by the NICE SCIE guideline. In addition, only the Italian guideline recommended memory training. On the basis of the same but one studies, the NICE SCIE guideline did not recommend memory training because of a lack of benefits beyond the particular tasks trained. The German guideline was the only guideline that recommended the use of validation.
The Spanish guideline mentioned several psychosocial interventions, but formulated no recommendations for about half of them. The psychosocial interventions that the Spanish guideline recommended were also recommended by the NICE SCIE guideline.
The two Dutch guidelines combined paid less attention to psychosocial interventions than each of the other five guidelines separately. The Dutch guidelines were the only ones that recommended emotion-oriented care and had the least in common with recommendations of the NICE SCIE guideline.
Five guidelines recommended interventions for family carers of people with dementia (Table 2). The NICE SCIE and the Italian guideline based their recommendations on more than 20 studies, which were generally similar. The SIGN guideline referred to only three studies, while five of the 11 references included in the Dutch guideline were Dutch publications (Wind et al., 2003). None of the guidelines were very specific about which carer interventions should be used, except the NICE SCIE guideline. It recommended that carers have a care plan and that interventions be tailored to specific needs. It also suggested different types of interventions (e.g. skills training, social support and psychoeducation).
Six guidelines included recommendations for the use of psychosocial interventions to treat neuropsychiatric symptoms. Both the Dutch and German guidelines for formal carers in institutional care aimed exclusively at managing neuropsychiatric symptoms.
The German, Spanish and NICE SCIE guidelines included general recommendations for the use of psychosocial interventions to treat neuropsychiatric symptoms. They summarised a range of interventions that might be effective because positive, though inconclusive, results were found for their use. In addition, the German and NICE SCIE guidelines included recommendations for assessing neuropsychiatric symptoms.
The Dutch NIVEL guideline and the Italian guideline linked specific interventions to specific neuropsychiatric symptoms. The first one recommended snoezelen for apathy in the later stages of dementia. The Italian guideline recommended aromatherapy for reducing the agitation and aggression of some patients. The SIGN guideline reviewed the same evidence base, but it concluded that aromatherapy could not be recommended for reducing specific neuropsychiatric symptoms. It also advised against multisensory stimulation for people with moderate to severe dementia.
Our inventory of dementia guidelines showed that recommendations for psychosocial interventions are included in dementia guidelines in five of 12 European countries. We compared the sections on psychosocial interventions in seven guidelines from these countries and found a wide variety between guidelines in the methodological quality and inclusion of recommendations for psychosocial interventions. The NICE SCIE guideline from the UK had the best methodological quality overall, and it included the most recommendations on psychosocial interventions.
Our study revealed important factors that were related to the inclusion of recommendations for psychosocial interventions in guidelines. These factors were the influence of guideline developers, the methodological quality of the guidelines and the appreciation of available evidence. The influence of guideline developers was clearly visible in one Dutch guideline that specifically stated that psychosocial interventions were not included because they involve nonmedical disciplines that were not present in the developing work group (van Gool et al., 2006). In contrast, another Dutch guideline and a German guideline were developed for formal carers in institutional care and included psychosocial interventions only (NIVEL, 2004a; NIVEL, 2004b; Bartholomeyczik et al., 2007). The guidelines that had better methodological quality, and especially, higher scores for the rigour of development domain, included more recommendations for psychosocial interventions in general (Table 2). Evidence was not appreciated in the same way across the guidelines. In the case of specific psychosocial interventions, such as aromatherapy or multisensory stimulation, review of the same evidence base resulted in disagreement between guidelines about whether to recommend it. Scientific evidence shows that psychosocial interventions are effective in dementia care in general, though the evidence for specific psychosocial interventions is still mixed and limited (Livingston et al., 2005; Kverno et al., 2009; Hulme et al., 2010). Inconclusiveness of the evidence base was also why, in some countries like Denmark and France, recommendations for psychosocial interventions were found in consensus papers only and not in dementia guidelines (Vasse et al., 2008).
Including recommendations for psychosocial interventions in guidelines is one thing; the practical implementation is another. A dementia survey in six European countries has found that, at the time of diagnosis, physicians recommend this type of intervention far less often than pharmacological treatments (Rimmer et al., 2005). Furthermore, a survey in the UK found that poor access to nonpharmacological interventions was a major limiting factor for their use in the treatment of neuropsychiatric symptoms (Bishara et al., 2009). Guidelines should encourage the implementation of recommendations by addressing the organisational barriers and the cost implications for applying them, but in our study, most guidelines had low scores for the AGREE applicability domain.
Most Europeans with dementia are cared for at home, and they want to stay at home as long as possible (Alzheimer Europe, 2006). In southern European countries, family members taking care of their ill relatives is a cultural tradition (Alzheimer Europe, 2006). Dementia care services (like home care, respite care and long-term care facilities) are lacking or unevenly distributed in these countries (Alzheimer Europe, 2006; Alzheimer Europe, 2007). Dementia patients often receive informal care only and the carer's burden in southern countries is the greatest in Europe (Alzheimer Europe, 2006; Hausner et al., 2010). Considering this, one might expect that guidelines from southern European countries would give extensive attention to carer support. Although the Spanish and Italian guidelines that we studied recommended carer interventions, the recommendations were rather general, and applicability in practice was barely addressed. One reason for this could be that both guidelines were national initiatives, whereas dementia care is organised on a regional and local level with great variability in services found across both these countries (Alzheimer Europe, 2007; Mateos et al., 2010). In contrast, the NICE SCIE and SIGN guidelines from the UK were national guidelines that did consider local implementation and gave directions for adaptation to the local situation.
There were national guidelines in Germany and the Netherlands as well, but most were aimed at specific professional groups or specific dementia-related issues, such as the use of restraints (Vasse et al., 2008). Only a few recommendations for psychosocial interventions were included in the Dutch guidelines; this is probably due to poor methodological quality and focus on pharmacological interventions. Furthermore, guidelines for specific professional groups do not promote collaboration, although collaboration between professionals is known to improve the quality of dementia care in both primary care and the institutional setting (Callahan et al., 2006; Fossey et al., 2006). High-quality, multidisciplinary, dementia guidelines like those from the UK could improve collaboration between professionals and promote the inclusion of psychosocial interventions. They might help remedy the lack of multidisciplinary dementia guidelines that was identified across Europe earlier (Waldemar et al., 2007).
For the support of family carers and the management of neuropsychiatric symptoms, scientific evidence shows that psychosocial interventions are most effective when they are individualised and tailored to the specific needs of the patient and the carer (Sörensen et al., 2002; Brodaty et al., 2003; Fossey et al., 2006; O'Connor et al., 2009). There are guideline recommendations that simply state that some psychosocial interventions might be effective for carers or the treatment of neuropsychiatric symptoms. Such recommendations do not enable healthcare professionals to provide tailored and individualised care. We did not find clear recommendations for care plans, the assessment of carers, or the assessment of neuropsychiatric symptoms except for those in the NICE SCIE guideline. The knowledge that interventions are more effective when they are tailored and individualised has emerged in recent years. Some of the guidelines had been published before such evidence was available. Guidelines should be updated periodically to keep them up to date with scientific evidence. We note that the description of an updating process was lacking in five of the guidelines reviewed here. However, an update of the Dutch guideline for general practitioners is expected in 2011.
The use of the INTERDEM network enabled us to study the dementia guidelines from many countries. Unfortunately, it was not feasible to gather information about dementia guidelines from all the European countries, and some limitations should be considered in the interpretation of the findings. Our methodology for finding dementia guidelines across countries makes it likely that all the national guidelines were identified, but we may have missed some regional and local guidelines. In addition, our study represents the methodological quality and the recommendations of guidelines that were available at the time of the survey only. Dementia guidelines were being developed in at least one country (Sweden), and more countries may have since published dementia guidelines that include psychosocial interventions. A final limitation is the fact that we used the AGREE instrument, not to assess methodological quality of a guideline as a whole, but only for sections about psychosocial interventions. The instrument is sensitive to quality differences in clinical guidelines, but it is not certain to what degree the instrument allows comparison when only specific guideline sections are assessed (The AGREE Collaboration, 2003).
The inclusion of psychosocial interventions in dementia guidelines across Europe is limited, despite the growing evidence base for their effectiveness in dementia care.
Healthcare professionals and policy makers should ensure that high-quality dementia guidelines that include recommendations for psychosocial interventions become available in all European countries. Guideline developers should ensure that guidelines are kept up to date with the emerging evidence, and they should encourage the implementation of recommendations by addressing the organisational barriers and cost implications for application. Researchers should focus their studies not only on finding significant effects for specific psychosocial interventions, but also on the development and evaluation of strategies for implementing psychosocial interventions effectively in daily practice.
Conflict of interest
BW contributed to the UK NICE SCIE guideline. MVD contributed to the Dutch Practice Guidelines for Dementia, second revision [NHG-standaard Dementie M21].
Guidelines for psychosocial interventions in dementia care were found in five of 12 European countries.
Existing dementia guidelines should be updated regularly to ensure that healthcare professionals can provide evidence-based psychosocial care to people with dementia and their carers.
Guideline developers, dementia researchers, policy makers and healthcare professionals should pay special attention to the implementation of dementia guidelines for psychosocial interventions in the next few years.
This paper was produced within the framework of the Alzheimer Europe project ‘European Collaboration on Dementia. EuroCoDe’, which received financial support from the European Commission under its programme for community action in the field of public health (2003–2008). Neither the European Commission nor any person acting on its behalf is responsible for any use that might be made of the following information. The authors gratefully acknowledge the co-financing from the Fondation Médéric Alzheimer for the coordination of EuroCoDe.