Characteristics of dementia-friendly hospitals: an integrative review

Background/Objectives Dementia-friendly initiatives are becoming common in society, politics, and research, including health care. Regarding efforts to improve care for patients with dementia in hospitals, the term dementia-friendly hospital is being used increasingly. However, a theoretical understanding of this term and the underlying concept are missing. This integrative review aims to identify current descriptions of dementia-friendly hospitals and to analyze their characteristics. Methods An integrative review was conducted. The databases MEDLINE, CINAHL, PsycInfo, Cochrane Library, and additional resources were searched. Two reviewers independently screened publications for inclusion. We extracted data from the included publications and analyzed the descriptions of dementia-friendly hospitals using inductive content analysis in an iterative process. Results We identified 4191 records and included 34 publications on 17 descriptions of dementia-friendly hospitals. These were found in the context of practice projects (n = 8), recommendations (n = 6) and research (n = 3). Our analysis resulted in six characteristics of dementia-friendly hospitals. Characteristics related to the patients and their care are continuity, person-centeredness, consideration of phenomena within dementia and environment. Additional characteristics are valuing relatives and knowledge and expertise within the hospital. Conclusion Dementia-friendly hospitals are currently characterized more by healthcare practices and professional dementia experts than by the results of empirical studies. Additionally, the perspective of people with dementia is underrepresented in current descriptions. Accordingly, further research is needed that involves people with dementia in order to develop a theoretical understanding and suitable concepts of dementia-friendly hospitals, since their perspective is essential. Supplementary Information The online version contains supplementary material available at 10.1186/s12877-022-03103-6.


Background
Dementia-friendly initiatives are becoming more and more common in a variety of settings, such as dementiafriendly communities [1], prisons [2], the arts [3], environmental design [4] and healthcare [5] and targeted in national dementia plans [6,7]. The World Health Organization's global action plan set a target to establish at least one dementia-friendly initiative in 50% of countries by 2025 [8]. Dementia-friendly initiatives focus on people with dementia and the protection of their human rights, their integration into society and the reduction in stigma [8]. They aim to create a supportive, inclusive and empowering social and physical environment for people with dementia, their families and caregivers [8,9].
The term "dementia-friendly" is also increasingly used in international publications and national dementia strategies in the context of hospitals [10][11][12]. Hospitals are in general an unbefitting environment for people with Open Access *Correspondence: Christina.Manietta@dzne.de dementia because of functional care, processes, architecture, noise and the presence of strangers [13][14][15]. Hospitalization is a burden for this patient group and their relatives [13,14] and is associated with negative experiences for both of them [16][17][18]. Patients with dementia described their hospital experiences as feeling lost in the environment, being an outsider, boredom, and a lack of intersubjective relationships [18]. Additionally, a hospital stay is often associated with various adverse events and leads to poor outcomes for patients with dementia, such as a longer hospital stay, postoperative complications or moving into a nursing home [13].
To improve the care of patients with dementia in hospitals, various efforts have been made, including the creation of dementia-friendly hospitals (DFHs). However, the term DFH is used heterogeneously, and a theoretical understanding of DFHs as well as the underlying concept are lacking.
Scientific overviews of dementia-friendly discourses in policy documents [19] and dementia-friendly initiatives in research exist [9,20]; however, there is a lack of literature reviews focusing on DFHs.
To close this research gap, we conducted an integrative review with the objectives (a) to identify current descriptions of DFHs and (b) to analyze the characteristics of DFHs.

Methods
An integrative review according to Whittemore and Knafl [21] was conducted to identify a broad range of descriptions of DFHs from different types of literature and to analyze them comprehensively. An internal preestablished review protocol guided the process.

Literature search
We decided to include research and non-research publications based on our review objectives. Accordingly, we conducted a comprehensive literature search to identify both types of publications. For the research publications, a systematic literature search in the databases MED-LINE (via PubMed), CINAHL (via Ebsco), Cochrane Library and PsycInfo (via Ebsco) was performed (latest search in March 2022). We derived the search terms first from our review objectives and then from the results of an initial publication search. We used indexing words and free search terms, which we clustered according to the PICo scheme (population, phenomenon of interest, context) [22]. The search string was developed by the first reviewer (CM) and checked by the second reviewers (DP, MR) using Peer Review of Electronic Search Strategies (PRESS) [23]. (The MEDLINE search string is shown in Additional file 2). Owing to the non-research publications also included, which are mainly gray literature, further search strategies were conducted following Godin et al. [24]. Accordingly, two gray literature databases (OpenGrey, Grey Literature Report), two subject databases (GeroLit, CareLit), three search engines (Google, Google Scholar, Livivo) and targeted websites (e.g., the Alzheimer's Association) were searched. For this purpose, we created simpler search strings based on the MEDLINE search string or, if this was not applicable, we used a simple combination of the search terms. For a Google search, we conducted a systematic procedure [24] and defined various simple search strings in English and German (Additional file 3). In addition, we performed forward citation tracking for the research publications via Google Scholar and backward citation tracking for all publications included.

Publication selection
We included research and non-research publications containing descriptions of DFHs. The description had to provide content regarding what constitutes a DFH (e.g., contributions including concepts, several components and/or several characteristics of DFHs). Individual components (e.g., special care units) or dementia-friendly initiatives with a narrowed focus (e.g., dementia-friendly education, dementia-friendly environmental design) were only included if they were part of a concept of a DFH and a publication on the whole concept had been published. The term "dementia-friendly hospital" or similar terms such as "dementia-sensitive hospital" had to be used explicitly. Publications published in German and English from 2010 onwards were included. The detailed inclusion criteria are shown in Table 1.
Two reviewers (Reviewer 1: CM, Reviewer 2: DP, MR) independently screened first the titles and abstracts and then the full texts of the potentially relevant publications, against the inclusion criteria (Table 1) using Covidence [25]. The Google search screening process was carried out following Godin et al. [24]. To capture as many of the most relevant hits as possible but also a feasible number to screen, the title and the short description underneath the first 150 hits of each search string were screened by one reviewer (CM). Potentially relevant publications were bookmarked, recorded in an Excel spreadsheet and screened in full text by two reviewers (CM, DP) independently against the inclusion criteria. The search was recorded in an Excel spreadsheet. All conflicts in the screening process were solved together by the screening team (CM, DP, MR).

Data extraction
We merged several publications of the same study, recommendation or project and counted them as one description of a DFH because they contained essentially the same components of a DFH and differed in their level of detail or included supplementary information (e.g., on project evaluation, participants, funding). Therefore, we grouped the included publications into primary and additional publications. Publications with the more detailed description of a DFH were referred to as primary publications. Publications with less information on the DFH or only containing supplementary information were referred to as additional publications.
We created a standardized form to extract the following information from the included publications: general information on the publications (e.g., author, year, country, publication type, funding), information on the study, recommendation, or practice project (e.g., aim, design/ methods, participants, target group) and information on the phenomenon of interest (e.g., term, definition, development, target group, hospital, key components). The data were extracted by one reviewer (CM or AR) and checked by a second reviewer (CM or DP).

Data evaluation
The aims of our review were to identify and analyze the current descriptions of DFHs, regardless of the publication quality or type. Each description of a DFH was assumed to contribute to the understanding of DFHs. Since the quality of the article was not relevant for our review aims and would not have influenced our publication selection, analysis (e.g., weighting of studies and practice projects) or results, we did not perform a quality assessment of the included publications [21].

Data analysis
First, we used the extracted data for a detailed illustration of the included descriptions. Then, we analyzed the included descriptions of DFHs using inductive content analysis [26] in an iterative process to identify the characteristics of DFHs. The description of the primary publication was analyzed first and then the analysis was supplemented by the descriptions from the additional publications, if available. The inductive content analysis process was divided into open coding, creating categories and abstraction [26]. First, the descriptions of DFHs were read intensively, notes and headings were made and subcategories were created. Then similar subcategories were grouped into categories and these were grouped depending on their content to create main categories [26]. The analysis was carried out by two reviewers (CM, DP) independently. The categories of the two reviewers were regularly discussed and merged during the analysis process by the screening team (CM, DP, MR) until all descriptions had been analyzed. Abstraction at the level of the main categories was carried out together in the screening team. Finally, the category system was peer-checked, discussed and adapted by the entire review team (CM, DP, AR, CK, MR). For the data analysis, we used MAXQDA 2020.4.1 [27], and for the graphical presentation, SimpleMind [28].

Criteria Definition
Population ▪The target population of the phenomenon of interest had to have patients with a diagnosed dementia or symptoms of dementia (e.g., unspecified dementia), who were hospitalized ▪There were no restrictions regarding the type and severity of dementia

Phenomenon of interest
▪The phenomenon of interest was the DFH. DFHs had to be described in detail, i.e. we included only publications which provide content regarding what constitutes a DFH (e.g., contributions including concepts, several components and/or several characteristics) ▪Individual components (e.g., special care unit) or dementia-friendly initiatives with a narrowed focus (e.g., dementia-friendly education, dementia-friendly environmental design) were only included if they were part of a concept of a dementia-friendly hospital and a publication on the whole concept had been published ▪The term "dementia-friendly hospital" or similar terms such as "dementia-sensitive hospital" had to be used explicitly. The term "dementia capable" was excluded because of its more narrowed focus on the needs of people with dementia [6] Context ▪Included are publications whose phenomenon of interest related to the hospital, regardless of the level of care or ownership of the hospital ▪The focus of the phenomenon had to be on the hospital level and not exclusively on specific areas (e.g., emergency room) or ward (dementia-friendly ward) ▪Excluded were rehabilitation facilities, psychiatric specialty hospitals and facilities that were outsourced by the hospital (e.g., urgent care, outpatient, high-tech home care) Types of literature ▪Any type of research publications (i.e., studies of any design, reviews, qualification theses) were included ▪Non-research publications such as books, recommendations, guidelines, project reports, practice articles and policy documents were considered Others ▪Languages: German and English ▪Year: from 2010

Results
We identified a total of 4191 records via databases and other methods. After deduplication, 3470 titles and abstracts were screened for inclusion. Of these, 528 publications were screened in full text. Most publications were excluded in the full-text screening because the term DFH was used but not specified at all or described in detail (n = 189). Finally, we included 34 publications on 17 different descriptions of DFHs ( Fig. 1).
Most of the descriptions of DFHs reflected the perspective of healthcare professionals or professional experts in the field (n = 14) [29-39, 41, 42, 45, 46]. Two of the descriptions included the perspectives of people with dementia and relatives of people with dementia [43,44], while three of the descriptions included the perspective of representatives of people with dementia (e.g., the Alzheimer's Association) [43,46,47]. Additionally, some descriptions of DFHs used the SPACE principles [48] of the Royal College of Nursing [37,40,41,45] and one the palliative care concept of Silviahemmet [29] as an underlying concept.
A detailed illustration of all included descriptions is provided in Table 2.

Characteristics of DFHs
We identified six characteristics of DFHs based on the analysis of the included descriptions: continuity, personcenteredness, consideration of phenomena within dementia, environment, valuing relatives and knowledge and expertise (the full description is presented in Additional file 4). Figure 2 shows an overview of all characteristics.

Person-centeredness
Person-centeredness in a DFH is characterized by the following aspects: knowing the person, attitude toward the person and caring for the person with dementia in a person-centered way. Knowing or getting to know the person with dementia means not only knowing the usual information collected in the hospital, such as medical history, but also the person beyond that. To get to know the person, information such as their behavior [32,33,38,42,54], their habits [32, 33, 36-39, 42, 50], their preferences [33,[36][37][38][39][40][41][42], their biography [31, 32, 38, 39, 41-43, 45, 46, 54] and their relatives and social circumstances [37,39,42,43] are needed. However, there are different views on this kind of information. In the study by Toubol et al. [43], participants with dementia and relatives of people with dementia believe that general information (e.g., current condition, how to fulfill their needs) is sufficient and that there is no need to know everything about the patient's personal life. However, biographical information is seen by professionals as important for conversations, relationship building and tailored care [32,39,[41][42][43].

Knowledge and expertise
In a DFH, both dementia-specific and multiprofessional knowledge and expertise are available. Dementia-specific knowledge and expertise are available at a basic level for all hospital staff [30, 31, 35-43, 45, 46]. In addition, there are dementia or geriatric experts (at ward or hospital level) who can be involved in the care of patients with dementia and support the health care staff [30][31][32][33][34][35][36][37][38][39][40][41][42]45]. Moreover, there is multiprofessional knowledge and expertise available for the care of patients with dementia. Therefore, professionals from diverse disciplines are involved in care, and find different ways of working together, such as holding multiprofessional team meetings or case conferences, to bundle their expertise and knowledge [31-42, 45, 46, 49, 53].

Characteristics related to the context of the description
All of the six characteristics of DFHs that we identified through our inductive content analysis were found in publications of practice projects, recommendations and research publications. All subcategories of the characteristics are essentially presented in their entirety in the publications of practice projects and recommendations. We identified fewer subcategories for all characteristics in two of the three included research publications [43,44]. In particular, the characteristic consideration of phenomena within dementia were not represented and only individual subcategories of the characteristics continuity and knowledge and expertise could be identified in these publications. These publications included the perspectives of people with dementia and their relatives, in contrast to the other publications which contained primarily the perspective of healthcare professionals or professional experts (Additional file 5 presents the characteristics in relation to each description).

Discussion
To our knowledge, our integrative review provides for the first time, a comprehensive overview of current descriptions of DFHs based on a systematic method and the inclusion of different types of literature. Moreover, our analysis clusters and synthesizes the characteristics of DFHs and presents them in a new, comprehensive manner. The term DFH was frequently used in all types of international literature, including research publications, but was rarely described. Even in the 17 descriptions included, only three provided a definition of the term, and some only described several components of DFHs without explicitly explaining what constitutes a DFH in its entirety. However, the described key components of DFHs did not differ substantially.
We found that the term DFH is more often described in publications of practice projects and recommendations than in research publications. Few of the descriptions were the result of an empirical study. Moreover, most of the included descriptions of DFHs were based solely on the perspective of healthcare professionals or other experts in this field. The perspective of people with dementia was rarely included, and only one description included different perspectives. Our results confirm that the involvement of people with dementia in the development of dementia-friendly initiatives is underrepresented [9]. However, to develop a concept of DFHs that is suitable for patients with dementia, it is essential to involve people with dementia in the development, since they are the only experts on what it feels like to live with dementia and only they can provide this unique perspective on the topic [55][56][57]. In particular, this would allow a differentiated view and a critical reflection on measures reported in the included descriptions, such as limited access to exits or the collection of personal information, as the study by Toubol et al. [43] partially shows. Nevertheless, it is important to involve different stakeholder groupspeople with dementia, their relatives, healthcare professionals and researchers -in the development of DFHs. This approach would lead to suitable concepts in a more differentiated and targeted way, and inappropriate interventions could be reduced.
We identified six characteristics of DFHs: continuity, person-centeredness, consideration of phenomena within dementia, environment, valuing relatives and knowledge and expertise. These characteristics address the needs of patients with dementia, their relatives and healthcare professionals as identified in other studies [15,16,[58][59][60][61]. In particular, person-centeredness is a central topic in other studies that interviewed patients with dementia and their relatives. Aspects, such as being accepted and treated as a valued person, being involved in their own care and conversations, and receiving personalized and tailored care, are described in these studies and seem to be important for this patient group [16,58,60,61]. In contrast to current initiatives for an age-friendly health system such as the 4Ms framework (What Matters, Medication, Mentation and Mobility) [62], which, among other things, considers the mentation of older people, the characteristics of a DFH we identified seem to address the needs of people with dementia more specifically and comprehensively.
In addition to the needs of people with dementia, their relatives and healthcare professionals, the characteristics we identified are also reflected in the goals of dementiafriendly initiatives in general. Accordingly, raising awareness about people with dementia, maintenance of their human rights, care that meets their needs and a supportive environment for them and their relatives [8] seems to be an overriding goal of DFHs. However, these goals can only be achieved by listening to and understanding people with dementia as "experts by experience" [55,57] and incorporating their perspectives into current healthcare practice and concepts of DFHs.

Limitations
There are potential limitations of our integrative review that need to be considered. We only included publications in English and German. This may have led to a bias, especially regarding the non-research literature, which is often published in the native languages. We conducted a comprehensive search of databases and used different strategies to identify non-research literature, but we might have missed publications or descriptions on the topic.
In addition, it should be noted that only publications that included a detailed description of DFHs were included; this may have biased the results. This review focused on DFHs; accordingly, only publications with this or similar terms (e.g., "dementia-sensitive") were considered to specify the understanding of DFHs. Publications without the term were excluded, which might have resulted in similar concepts described in other terms being overlooked.

Conclusion
Our integrative review provides an overview of current international descriptions of DFHs. Moreover, our analysis provides an initial understanding of DFHs and their characteristics and contribute to a consistent terminology of DFHs in the future. Although our identified characteristics are consistent with the needs of patients with dementia, it is important to keep in mind that these results are mainly derived from descriptions of DFHs based on the perspective of healthcare professionals and professional dementia experts. Accordingly, these characteristics can only be used as orientation for health care practice and research at this stage. However, our results can contribute to raising the awareness of healthcare professionals. Based on our identified characteristics of DFHs, healthcare professionals can reflect on their provision of care, hospital structures and processes and thus provide an impulse to move towards a DFH. For researchers, our results are important because they form an essential basis for further research in this field. Additionally, future research needs to explore synergies and overlaps between DFHs and other approaches such as age-friendly initiatives (e.g., 4Ms framework). This includes exploring whether these initiatives could be supplemented by aspects of DFHs. Based on our review results, future research needs to consider the following: (1) to reflect the characteristics of DFHs with people with dementia, their relatives and healthcare professionals, (2) to operationalize and tailor the characteristics of DFHs together with these target groups, and (3) to implement and evaluate the characteristics of DFHs across hospitals (nationally and internationally) so that country-specific and cross-country comparisons become possible. Since the perspective of people with dementia and their relatives on DFHs was underrepresented in the included descriptions, it is of utmost importance to involve them in a participatory way in the future.