Balancing individual rights and risks: a systematic review of qualitative studies of perspectives on older adults’ alcohol use in residential care settings

Abstract

An increasing number of older adults require residential care. Concurrently, older adults’ alcohol use is increasing. This review explored the perspectives of all relevant stakeholders on older adults’ alcohol use within residential care settings, through a systematic review and thematic synthesis of qualitative studies. Eight databases were searched for qualitative studies focusing on older adults’ alcohol consumption (defined as aged ≥ 50) within residential care settings, sampling any involved stakeholders, published up until January 2024. Quality appraisal utilised the Critical Appraisal Skills Programme checklist and included 15 studies of mainly moderate quality across seven high-income countries, reporting data from a range of stakeholders and representing varied older adults’ alcohol histories. Three themes were identified: alcohol use by older adults is socially acceptable and purposeful in residential care settings; alcohol helps in the pursuit of an ‘ideal’ outcome; and decision-making around older adults’ alcohol use varies depending on the involvement, knowledge, skills and beliefs of the participating stakeholders, who also vary. Reports of problematic alcohol use were rare and older adults in residential care settings should be supported to exercise their own choice in determining their alcohol use. However, residential care settings face particular challenges in managing the alcohol intake of older adults with limited mental capacity and alcohol dependency; owing to a lack of guidance, front-line staff make subjective decisions. Future research should develop guidance that involves all relevant stakeholders, including family members. Limitations include lack of generalisability to low- and middle-income countries and limited availability of raw data.

Introduction

Global population ageing is accelerating and expected to continue over the coming years (Rafalimanana and Lai 2013), driving a rise in demand for residential care (Hu et al. 2023) which describes anywhere older adults live outside of private residences (Douma et al. 2017) with the availability of assistance. Concurrently, in recent years there has been a small but steady increase in the amount of alcohol consumed by older adults (Bye and Moan 2020; John 2018). An inclusive definition of older adults includes all adults aged 50 and over (Bareham et al. 2019, 2020), as acknowledged by a UK charity for older adults (Age 2024). Data from 21 countries indicates that just over half (52%) of older adults drink alcohol each year, and of those, 12 per cent of men and women drink more than two or three drinks per day or report drinking more than four or five drinks on a single occasion, respectively (Calvo et al. 2020). Cultural perceptions of problematic alcohol use differ (Calvo et al. 2021), although excessive alcohol consumption is associated with serious social, psychological, physical and economic costs (Prime Minister’s Strategy Unit 2003). The evidence around alcohol use in older adults specifically can be mixed, with some literature showing the protective effect of moderate alcohol consumption (Corrao et al. 2000; Elkind et al. 2006; McCaul et al. 2010). However, a recent large-scale study identified an increasing risk of cancer, cardiovascular and all-cause mortality with increasing alcohol use, particularly in individuals of low socio-economic status (SES) and with existing health risk factors (Ortolá et al. 2024). The authors attributed this finding to the use as a reference point of occasional drinkers instead of abstainers, as the latter group often includes older adults with histories of high alcohol use and dependency, which inflates the health risks (Ortolá et al. 2024). Despite the health risks of alcohol use at any level in older adults, for many older adults alcohol use contributes to quality of life (Bareham et al. 2020; Kelly et al. 2018) through facilitating socialization (Bareham et al. 2019, 2020; Kelly et al. 2018; Parke et al. 2018) and providing a habitual pleasurable experience (Kelly et al. 2018). Thus, as an increasing number of older adults come to call residential care settings (RCSs) their home, it is relevant to consider both the value and the risks of their alcohol use within RCSs, how their alcohol consumption is managed and how relevant stakeholders are affected.

A qualitative methodology is appropriate for understanding older adults’ alcohol use in the context of RCSs (Godfrey 2015) as it involves a complex interplay of the thoughts, attitudes and actions of multiple stakeholders in the context of local and national policy landscapes (Bareham et al. 2019, 2020). Qualitative research facilitates understanding around a complex behaviour (alcohol use) and how it is influenced by its context (RCSs that vary in level of care provided, regulations and organizational culture) (Godfrey 2015). Thus, qualitative research is most appropriate for understanding how and why alcohol use differs across RCSs (Godfrey 2015). Despite this, few qualitative studies have examined alcohol use in RCSs, although existing studies indicate a symbolic value of rituals around alcohol (Emiliussen et al. 2021) that appears to be common across settings (Bareham et al. 2019). Systematic reviews of qualitative research have explored the experiences, attitudes and perceptions of both older adults (Bareham et al. 2019; Kelly et al. 2018) and health-care professionals (Bareham et al. 2020) across settings. However, available reviews provide no comparison between RCSs and private residences (Bareham et al. 2019). Indeed, primary studies indicate specific challenges associated with alcohol use within RCSs given that residents may be unable to access alcohol themselves and thus rely on others to provide it (Emiliussen et al. 2021); individual autonomy and needs must be balanced with communal living (de Graaf et al. 2022) and the provision of care provides opportunity to control older adults’ alcohol consumption (Emiliussen et al. 2021). Alcohol use within RCSs is also situated within a wider issue of the right to autonomy as conflicting with living in RCSs when older adults are not able to participate in decisions that concern them (Hedman et al. 2019). The RCSs within primary studies vary widely on their residents and approach to alcohol use; thus, a synthesis is needed to understand what works and for whom. Furthermore, existing systematic reviews (Bareham et al. 2019, 2020; Kelly et al. 2018) excluded studies of older adults with alcohol dependency, meaning that it is uncertain how alcohol dependency is managed in RCSs. In summary, available reviews provide a useful overview on the breadth of research into alcohol use in older adults. However, focused reviews are also needed to provide in-depth comparison and understanding of which stakeholders are involved in older adults’ alcohol consumption within RCSs, how experiences differ across stakeholders and the unique challenges around management of residents’ alcohol use. There also remains a need to adopt a more pragmatic and rights-based perspective that acknowledges and accepts alcohol, given that abstinence may not be possible or feasible (Nixon and Burns 2022). The authors have previously published a corresponding in-depth and pragmatic review on alcohol use within older adults who receive domiciliary care (Haighton et al. 2024), facilitating comparisons to be made across settings. Thus, the current study aimed to examine the perspectives of relevant stakeholders on alcohol use in RCSs by older adults.

Methods

This study describes a systematic review and thematic synthesis of qualitative studies with the protocol pre-registered via PROSPERO (registration number: CRD42024504197) and following the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines for reporting of systematic reviews (Page et al. 2021) (see Supplementary Material 1). The review question was defined using the SPIDER framework (Sample, Phenomenon of Interest, Design, Evaluation, Research type) to synthesise all available qualitative research exploring older adults’ alcohol use within RCSs, through including the views of all relevant stakeholders involved in supplying it. Specific aims included to explore the similarities and differences between views of different stakeholders (e.g. older adults, staff and family) and according to differing alcohol use across RCSs (e.g. nursing homes, care homes and sheltered accommodation). Prior to pre-registration, PROSPERO, Joanna Briggs Institute Registries and Open Science Framework were searched for reviews of a similar scope, of which none were identified.

Search strategy and selection criteria

The SPIDER framework for qualitative reviews was used to inform eligibility criteria (Cooke et al. 2012). The focus was on older adults’ alcohol consumption within RCSs. Although the definition of ‘older adult’ varies widely and is relatively arbitrary, 50 and over was selected as the definition of older adults to be inclusive to relevant studies and facilitate comparison across settings with existing reviews (Bareham et al. 2019, 2020). Where some participants were younger than 50, studies were included if the mean age was 50 or above. Older adults with any drinking history (from occasional use to alcohol dependence) were included, although studies where all participants were abstinent (e.g. in the case of a blanket ban on alcohol [Wadd et al. 2024]) were excluded given that the aim of the study was around management of alcohol use. Samples could include any stakeholders involved in older adults’ alcohol use in RCSs, including staff and family members. We defined RCSs, for the purpose of this review, as anywhere that care is delivered for a prolonged period of time (including assisted living facilities, care homes and continuing care retirement communities); thus, private dwellings were excluded. The inclusion of sheltered accommodation in the definition of RCSs differs from the updated Census definition (Office for National Statistics 2021) as there is a constant assurance of support and supervision if needed. Where studies included a mixture of substances or types of setting, studies were included only if quotes were separable by setting or substance. Only qualitative studies were included to explore perceptions and attitudes in depth; thus, quantitative studies and reviews were excluded. Where studies employed mixed methods, qualitative data were extracted where possible. Finally, inclusion criteria for research type included articles in English written in any year that were peer reviewed or grey literature, whilst book reviews, conference abstracts, editorials, opinion pieces and commentaries were excluded. Owing to a lack of resources for translation services, only articles published in English were included.

Eight databases were comprehensively searched on 21 January 2024 for eligible studies (Medline, ASSIA, APA PsycArticles, Nursing and Allied Health Database, Psychology Database, Public Health Database, and ProQuest Dissertations and Theses Global via ProQuest and CINAHL via EBSCO). The search strategy is displayed in Table 1 and the specific searches applied for ProQuest and EBSCO are supplied in Supplementary Material 2. The search as applied to title and abstract was based on the sample (older adults), the phenomenon of interest (alcohol use), the setting (residential care) and the design (qualitative), and included truncations, wildcards and limits (humans and English language) as appropriate.

Table 1. Search strategy (the specific search for each search engine is supplied in supplementary material 1)

Study selection

Following the search, references were exported to Endnote and duplicates removed. The remaining articles were uploaded to Rayyan for screening based against the inclusion criteria; firstly based on title and abstract and the remaining articles on full text, with reasons for exclusion at this stage recorded. Screening was conducted independently by two reviewers (BN and CC) and any disagreements were resolved through discussion with a third independent reviewer (CH). Inter-rater reliability was calculated at each stage using Cohen’s kappa (McHugh 2012), applying the conservative parameters set by Altman (1990). Inter-rater reliability for screening of title and abstract was good (k = 0.633) and very good for full text, of which there was 100 per cent agreement (k = 1.000). To facilitate open science practices, both title and abstract and full text screening stages are publicly available via Rayyan: (https://rayyan.ai/reviews/908744 and https://rayyan.ai/reviews/916540, respectively).

Data extraction and quality appraisal

Data extraction was conducted by two independent reviewers (BN and CC) and followed a standardized data extraction form that recorded data related to study characteristics, aims, context, methodology, participant details, headline findings, number of quotes and author conclusions. The bespoke data extraction form was developed in accordance with the specific aims of the current review, using guidance from Cochrane (Cochrane Effective Practice Organisation of Care 2017; Noyes et al. 2018). All direct quotations were extracted from each paper aside from one to two-word quotes where no context was provided. Themes relevant to the research question and their description and explanation were also extracted. To facilitate comparison between included RCSs, the settings were categorized according to their resource utilization using the Resource Utilization Groups 4th version (RUG-IV) (Fries et al. 1994); more information is provided later in the ‘Data analysis’ section. Thus, the data extraction form also recorded both information on whether RUG was explicitly cited and details of the setting that would help inform categorization. Again, any disagreements were resolved through discussion with a third reviewer (CH). In accordance with open science practices, the completed data extraction Excel file is publicly available via OSF: https://osf.io/e4pmt/?view_only=2702d2a3da674972b7b30041f29664ed.

Quality appraisal was conducted by two independent reviewers (BN and CH) by applying to the included studies the Critical Appraisal Skills Programme (CASP) quality appraisal tool for qualitative studies (CASP 2018). To facilitate a comprehensive overview of the existing literature, studies were not excluded based on quality but rather informed the interpretation of findings. To facilitate the meaningful incorporation of study quality, acknowledged as an important component of qualitative evidence synthesis, two CASP items were selected as the most relevant for the current review’s aims to inform the overall quality judgement of the included studies. Included studies were scored according to adherence to items five (‘were the data collected in a way that addressed the research issue?’) and eight (‘was the data analysis sufficiently rigorous?’). Specifically, in accordance with Cochrane guidance (Noyes et al. 2018) and in the absence of evidence-based standardized criteria (Long et al. 2020), each study was categorized into ‘low’ (neither item was addressed), ‘medium’ (one item was addressed) or ‘high’ (both items were addressed) quality. Again, inter-rater reliability was calculated using Cohen’s kappa and the conservative parameters set by Altman (1990). Any disagreements related to quality appraisal were resolved through discussion between both reviewers. Inter-rater reliability for quality appraisal was moderate (k = 0.536), and there was 80 per cent agreement between raters.

Data analysis

Included studies were classified according to the Resource Utilization Groups 4th version (RUG-VI) classification system, which separates residential settings into seven main categories according to the level and specialty of care required (North Dakota Department of Human Services 2019). Within each category, settings are further classified according to the level of assistance with activities of daily living (ADLs) and the presence of restorative nursing. Where multiple contexts were included, studies were classified according to the most specialized care (the highest classification). Classification was based on study context rather than on the participants. Where information was lacking, classification was inferred by the type of residential setting included. Assistance with ADLs was not assumed unless ADLs were specifically mentioned. Where ADLs were mentioned but not specified, the average of the ADL scale was selected to inform classification. Classification was conducted by the primary researcher (BN) and checked by another author experienced in applying the RUG-VI (JS).

All extracted data, including quotes and themes and their descriptions, were uploaded to Nvivo (Lumivero 2023) for analysis. Line by line free coding was applied to both sets of data (direct quotations and interpretations by the authors of the included studies). Initially, codes were mainly descriptive; then they were built into analytical themes. Codes reflected second-and third-order constructs described in meta-ethnography: codes that were described by authors of included studies, and codes identified by the review team to describe patterns and differences across included studies, respectively (Noblit and Hare 1988). In accordance with a reflexive approach to thematic analysis (Braun and Clarke 2019), independent coding was not conducted (Braun and Clarke 2021). Instead, all analysis was conducted by the primary reviewer (BN) and meetings were held with others in the team (CH and CC) throughout the data analysis to reflect upon and refine the themes and sub-themes. Furthermore, the primary researcher (BN) recorded reflective notes throughout the data analysis.

Results

Characteristics of included studies

The PRISMA diagram to illustrate study selection is shown in Figure 1 and reasons for exclusion of studies based on full text screening are displayed in Supplementary Material 3. Fifteen studies were included, representing 488 participants (accounting for four overlapping participants across two included studies [de Graaf et al. 2022; de Graaf et al. 2023b]), not including staff focus groups of unspecified size (McCann et al. 2017) or participants that did not contribute to relevant themes (David et al. 2023; Philpin et al. 2011). Table 2 provides a summary of included studies. Included studies were published between 2002 (Klein and Jess 2002) and 2023 (David et al. 2023; de Graaf et al. 2023b), possessed sample sizes ranging from 5 (Pollak 2016) to 197 (Wadd et al. 2024) and mostly represented the USA (Burruss et al. 2015; Chambers 2020; David et al. 2023; Klein and Jess 2002; Pollak 2016), followed by England (McCann et al. 2017; Payne 2018; Wadd et al. 2024), Norway (Johannessen et al. 2021; McCann et al. 2017) and the Netherlands (de Graaf et al. 2022; de Graaf et al. 2023b), then Denmark (Emiliussen et al. 2021), Australia (Dare et al. 2014) and Canada (Nixon and Burns 2022).

Figure 1. PRISMA diagram to depict the search and selection process (Page et al. 2021).

Table 2. Summary table of the characteristics and relevant themes identified by each included study

Notes: ‘Sample’ column is displayed as: sample size (type of participants and composition of the sample), inclusion criteria for participants, age range of older adults (mean [M] and standard deviation [SD] if provided) and drinking habits of older adults. Themes were identified by the authors of the included primary studies, although only themes relevant to the aims of the current review were extracted.

Settings ranged from independent living such as ‘sheltered accommodation’ (Burruss et al. 2015; Dare et al. 2014; Payne 2018) to care facilities providing care from health-care professionals including nursing homes (de Graaf et al. 2022; de Graaf et al. 2023b; Johannessen et al. 2021; Pollak 2016). According to the RUG-IV classification, most included settings cared for residents with reduced physical function, with (Chambers 2020; David et al. 2023; McCann et al. 2017) or without (Burruss et al. 2015; Dare et al. 2014; Emiliussen et al. 2021; Johannessen et al. 2021; Klein and Jess 2002; Nixon and Burns 2022; Payne 2018; Philpin et al. 2011) support with ADLs. A minority of studies were classified to care for residents with behavioural symptoms or reduced cognitive performance as they provided care for residents with dementia (de Graaf et al. 2022; de Graaf et al. 2023b; Wadd et al. 2024), and another study provided rehabilitation within a nursing home (Pollak 2016). In the studies where it was clear, all but one (Dare et al. 2014) RCSs provided meals for residents (Burruss et al. 2015; Chambers 2020; David et al. 2023; Emiliussen et al. 2021; Johannessen et al. 2021; Klein and Jess 2002; McCann et al. 2017; Nixon and Burns 2022; Philpin et al. 2011; Pollak 2016; Wadd et al. 2024). For other settings, it was unclear if meals were provided or if residents catered for themselves (de Graaf 2022; de Graaf et al. 2023b; Payne 2018), and it was often uncertain whether residents could independently shop for groceries to supplement provided meals (and therefore were able to buy alcohol independently).

All studies utilized interviews and some also conducted focus groups with specific stakeholder groups (Johannessen et al. 2021; McCann et al. 2017; Philpin et al. 2011; Wadd et al. 2024) or combined discussion with participant observations (David et al. 2023) alone or with document analysis (McCann et al. 2017; Philpin et al. 2011). Most often, included studies sampled residents (Burruss et al. 2015; Dare et al. 2014; David et al. 2023; de Graaf et al. 2022; de Graaf et al. 2023b; Emiliussen et al. 2021; McCann et al. 2017; Nixon and Burns 2022; Payne 2018; Wadd et al. 2024) and staff (de Graaf et al. 2023b; Emiliussen et al. 2021; Johannessen et al. 2021; Klein and Jess 2002; McCann et al. 2017; Nixon and Burns 2022; Philpin et al. 2011; Pollak 2016; Wadd et al. 2024), although a minority of studies included family or partners (Chambers 2020; de Graaf et al. 2023b; Emiliussen et al. 2021), setting managers (Emiliussen et al. 2021; McCann et al. 2017; Wadd et al. 2024) or inspectors (Wadd et al. 2024). Of those that included staff, studies most commonly sampled care professionals (de Graaf et al. 2023b; Emiliussen et al. 2021), although social workers (Pollak 2016), registered nurses (Johannessen et al. 2021) and general practitioners (McCann et al. 2017) were also represented. Five included studies adopted a holistic perspective and sampled a range of relevant stakeholders (de Graaf et al. 2023b; Emiliussen et al. 2021; McCann et al. 2017; Nixon and Burns 2022; Wadd et al. 2024), including between two (Nixon and Burns 2022) and four (Emiliussen et al. 2021; Wadd et al. 2024) different stakeholder groups. The available literature within care and nursing homes mostly focused on staff perspectives (Johannessen et al. 2021; Philpin et al. 2011; Pollak 2016) or adopted a holistic approach by including multiple stakeholders (de Graaf et al. 2023b; Emiliussen et al. 2021; McCann et al. 2017; Wadd et al. 2024), whilst studies of sheltered accommodation almost wholly focused on residents’ perspectives (Burruss et al. 2015; Dare et al. 2014; Nixon and Burns 2022; Payne 2018), and none included staff perspectives. Similarly, the two studies focused on assisted living included only residents (David et al. 2023) and family members (Chambers 2020) and did not include those involved in the management of RCSs or in delivering care.

Seven included studies were judged to be medium quality (Burruss et al. 2015; Chambers 2020; Dare et al. 2014; de Graaf et al. 2023b; Emiliussen et al. 2021; Johannessen et al. 2021; McCann et al. 2017), whilst four studies each were judged to be of high (David et al. 2023; de Graaf et al. 2022; Nixon and Burns 2022; Payne 2018) or low (Klein and Jess 2002; Philpin et al. 2011; Pollak 2016; Wadd et al. 2024) quality (see Supplementary Material 4 for scoring of specific items).

Themes

The overall coding framework, with sub-themes, example codes and illustrative quotes, is provided in Table 3. The three themes are first discussed, followed by an overall comparison across RCSs.

Table 3. Coding framework with example individual codes and quotes

Theme 1: Alcohol use by older adults is socially acceptable and purposeful in residential care settings

The variation of alcohol consumption observed across RCSs is comparable to the general population and is merely continued into RCSs. Alcohol consumption varied across residents from abstinence (Emiliussen et al. 2021) to dependency (McCann et al. 2017; Nixon and Burns 2022). Mostly, consumption was low to moderate (Emiliussen et al. 2021; Johannessen et al. 2021), and problematic use (an immediate negative impact on themselves or others) was rare (Chambers 2020; Emiliussen et al. 2021; Johannessen et al. 2021; Klein and Jess 2002; McCann et al. 2017; Pollak 2016). Consumption either remained the same (Chambers 2020) or decreased (Chambers 2020; McCann et al. 2017; Nixon and Burns 2022) on admission to the RCS. It is noteworthy that some studies reported that RCSs did not admit older adults with alcohol dependency (Klein and Jess 2002; Wadd et al. 2024), potentially explaining the low rates of problematic consumption reported. Two studies described ‘wet’ (McCann et al. 2017) RCSs that cared for individuals with alcohol addiction and dependency (McCann et al. 2017; Nixon and Burns 2022) and reported that, independent of a reduction in consumption on admission, the impact of residents’ consumption reduced as residents received care for their basic needs and accessed primary care (Nixon and Burns 2022).

Again, independent of the RCSs and their policies relating to alcohol use, alcohol use by older adults was facilitated or reduced depending on a range of factors comparable across the general population. Independent of admission to RCSs, alcohol use could increase owing to contextual factors such as grief (Chambers 2020; Payne 2018) and mental health difficulties (Chambers 2020; Payne 2018), or decrease owing to factors including driving (Dare et al. 2014), older age (de Graaf et al. 2022), health issues (Emiliussen et al. 2021) and medication use (de Graaf et al. 2023b; Payne 2018). Retirement could either increase or decrease alcohol use through a decrease in responsibilities (Dare et al. 2014) or being linked to work (Dare et al. 2014; Payne 2018). Similarly, alcohol use could be increased during socialization as older adults were less aware of their consumption and were encouraged by social norms around drinking alcohol amongst peers (Burruss et al. 2015; Dare et al. 2014; Payne 2018), or could instead be increased when alone as some individuals felt shame or worry towards what others would think about their consumption (Chambers 2020; Dare et al. 2014; Payne 2018). The three studies describing independent living mainly focused on influences independent of the setting, including driving, habits and routine; the only influence of the setting discussed was facilitatory owing to an increase in social activities (Burruss et al. 2015; Dare et al. 2014). However, there was less engagement with social activities in RCSs that housed residents with a wide age range (McCann et al. 2017; Payne 2018), which was particularly a concern where social activities aimed to divert attention away from alcohol (McCann et al. 2017). The RCSs varied widely on the facilitation of alcohol use, ranging from a blanket ban to bars on site with no rules around alcohol (Klein and Jess 2002; Wadd et al. 2024). However, endorsement of alcohol use by RCSs seemed to be independent of problematic use as residents could find ways of accessing more alcohol than was allowed (McCann et al. 2017).

Similarly, on an individual level, alcohol use was perceived to help fulfil the basic psychological human needs of autonomy (the freedom of choice and actions aligning with one’s sense of self) and relatedness (connection to others) (Deci and Ryan 2000). Related to autonomy, alcohol use was perceived to encourage choice and autonomy (Emiliussen et al. 2021; Johannessen et al. 2021), routine and continuation of habits (Burruss et al. 2015; Dare et al. 2014; de Graaf et al. 2022; Payne 2018) and possessing a stable self-identity, which were perhaps hindered on admission into RCSs. Staff and family were reluctant to remove another choice from residents’ lives owing to an empathy that their ability to choose was already limited within RCSs. However, this respect was contrasted with the dependence on family and staff, particularly for residents with limited physical functioning compared to other residents, including residents with dementia (de Graaf et al. 2022), to supply alcohol, which meant that the choice of family and staff was inevitably involved. Similarly, there was an overall respect for agency within RCSs by all stakeholders which extended to alcohol use. Removal of choice around alcohol was a barrier to feeling a sense of home and belonging (McCann et al. 2017). A desire for a ‘normal’ (Philpin et al. 2011; Wadd et al. 2024) home environment incorporated both the ability of alcohol to provide a feeling of home, ‘hygge’ or being ‘cosy’ (Emiliussen et al. 2021) within RCSs (Emiliussen et al. 2021; Nixon and Burns 2022) and a stable constant throughout the transition to the RCS (Wadd et al. 2024) and changing friendships, retirement, loss and a reduction in physical and/ or psychological functioning (Dare et al. 2014). Related to connectedness, alcohol use fostered a sense of feeling at home and belonging (Emiliussen et al. 2021).

Additionally, a wider societal context of social acceptability of alcohol was continued into RCSs (de Graaf et al. 2022). Alcohol was generally viewed positively by all stakeholders, as a ‘treat’ (Dare et al. 2014) or as symbolic of celebration (Emiliussen et al. 2021; Payne 2018; Philpin et al. 2011) and positive memories (Payne 2018), and thus inherently tied to mealtimes (Burruss et al. 2015; Philpin et al. 2011) or specific occasions or rituals (Emiliussen et al. 2021; Payne 2018; Philpin et al. 2011). Generally, residents did not feel that they had a problematic relationship with alcohol as drinking was normalized amongst their peers (Burruss et al. 2015; Dare et al. 2014; Emiliussen et al. 2021). Positive attitudes towards alcohol use in RCSs were contrasted with more negative and paternalistic attitudes towards smoking (de Graaf et al. 2022; de Graaf et al. 2023b), which was attributed to a wider societal shift towards negative perceptions of smoking (de Graaf et al. 2022), a lower prevalence of smoking amongst participants (de Graaf et al. 2022; de Graaf et al. 2023b) and a view that smoking disrupts others more than low-level drinking (de Graaf et al. 2022), as alcohol use by residents was no longer accepted if it affected others (de Graaf et al. 2022; de Graaf et al. 2023b) or led to addiction or dependency (McCann et al. 2017; Nixon and Burns 2022). Generally, there was a sense amongst family members (Emiliussen et al. 2021; Wadd et al. 2024) and staff (de Graaf et al. 2023b; Emiliussen et al. 2021; Klein and Jess 2002; Wadd et al. 2024) that older adults were entitled to access alcohol and that it formed a fundamental part of their quality of life (de Graaf et al. 2023b; Emiliussen et al. 2021; Johannessen et al. 2021; Wadd et al. 2024). Facilitation of alcohol use by RCSs (Burruss et al. 2015; Chambers 2020; Dare et al. 2014; Emiliussen et al. 2021; Johannessen et al. 2021), including happy hours (Chambers 2020; Dare et al. 2014) and a pub on site (Burruss et al. 2015), tended to be popular and accepted by all stakeholders. Interestingly, a socio-economic gradient in acceptability and facilitation of alcohol was notable whereby alcohol use of residents of higher SES seemed to be more acceptable to staff (Pollak 2016; Wadd et al. 2024) and facilitation of alcohol use by setting was higher in RCSs that catered to more affluent residents owing to higher resources and staff capacity (Wadd et al. 2024). Harm reduction was also recognized as safer and more inclusive for those with alcohol dependency, and a more realistic goal than abstinence was in an absence of alcohol-related problems (McCann et al. 2017).

Within wet RCSs specifically (the only setting-specific observation within this theme), the acceptance and inclusivity provided fostered a sense of belonging for residents with alcohol dependency who often face exclusion from RCSs (McCann et al. 2017; Nixon and Burns 2022). Particularly, a harm reduction policy, the acceptance of alcohol use with efforts to manage and minimize the risks associated with it, fostered a sense of acceptance, which increased satisfaction amongst residents and trusting relationships with staff (Nixon and Burns 2022). When trust was established, staff were able to enforce some control over residents to reduce the harmful effects of alcohol whilst still retaining trust and cooperation with the residents (Nixon and Burns 2022).

Theme 2: Alcohol helps in the pursuit of an ‘ideal’ outcome

Generally, it was recognized by all stakeholders that alcohol helped to achieve positive outcomes for most residents across RCSs (Burruss et al. 2015; Chambers 2020; Dare et al. 2014; de Graaf et al. 2022; Emiliussen et al. 2021; Johannessen et al. 2021; Wadd et al. 2024). Residents described outcomes including a ‘social lubricant’ (Dare et al. 2014) and a sense of togetherness (Burruss et al. 2015; Dare et al. 2014; Emiliussen et al. 2021; Philpin et al. 2011), an emotional response such as calm (Burruss et al. 2015) or relaxation (Burruss et al. 2015; Dare et al. 2014) and the continuation of routine and opportunity to exercise choice, as discussed earlier. However, a small number of studies suggested that this idealized version of intended outcomes of alcohol use was more individualized, as some residents demonstrated negative outcomes including antisocial behaviour (although this was rare) (Chambers 2020) and deterioration of health such as encouraging falls (Chambers 2020) or decreasing the efficacy of medication (Klein and Jess 2002). Impacts on others were more salient to residents than negative impacts on their own health. Also, staff across care and nursing homes reported that non-alcoholic alternatives could also help to achieve some of the idealized outcomes of alcohol without any negative effects including toasting (Emiliussen et al. 2021) or feelings of togetherness (a feeling of closeness or unity with others) (Emiliussen et al. 2021; Johannessen et al. 2021). Nonetheless, non-alcoholic alternatives were mainly viewed by staff to accompany rather than replace alcohol use, for example after residents had already consumed alcohol (Emiliussen et al. 2021). Furthermore, the use of non-alcoholic beverages or ‘watered down’ alcohol by staff was not always with the consent of residents with alcohol dependency or dementia (Pollak 2016). Residents did not discuss use of non-alcoholic beverages to manage their own intake, and this sub-theme was not relevant across sheltered accommodation and assisted living RCSs.

Theme 3: Decision-making around older adults’ alcohol use varies depending on the involvement, knowledge, skills and beliefs of the participating stakeholders

A running theme across RCSs was the need to balance older adults’ right to choose to use alcohol based on their needs and preferences, and the risks to themselves and others. Within sheltered accommodation, concerns about risk were mainly around ensuring resident safety and retaining their housing (Nixon and Burns 2022), whereas the risks discussed in care and nursing homes were often around the impact on other residents and a need to enforce protective policy (Emiliussen et al. 2021; Wadd et al. 2024). For example, serving alcohol on special occasions was perceived as balancing older adults’ wants with their health limitations (Emiliussen et al. 2021).

Interestingly, there was no sense of enforcement or management of alcohol use of older adults by staff or family members within the data from the settings of most independence (e.g. sheltered accommodation and retirement villages) (Burruss et al. 2015; Dare et al. 2014; Payne 2018). Instead, the only impositions on older adults’ alcohol use were the limitations set on their own use (Dare et al. 2014) or through social norms (Dare et al. 2014). Residents self-managed their own intake or, within wet care homes (Dare et al. 2014), were supported and encouraged to self-manage (Nixon and Burns 2022). Whilst this may be owing to a difference in study aims, where studies aimed to investigate factors influencing alcohol use and the subsequent involvement of residents only (Burruss et al. 2015; Dare et al. 2014; Payne 2018), rather than the involvement or experiences of other stakeholders in relation to older adults’ alcohol use (de Graaf et al. 2022, 2023b; Emiliussen et al. 2021; Johannessen et al. 2021; Klein and Jess 2002; McCann et al. 2017; Nixon and Burns 2022; Pollak 2016; Wadd et al. 2024), this does indicate that issues of decision-making around alcohol use are present only when individuals lack some functional or mental capacity. Common across all settings was a distinct lack of specialized training for staff (Johannessen et al. 2021; Klein and Jess 2002; Wadd et al. 2024), although only staff from care and nursing homes expressed a need for guidance and training (Johannessen et al. 2021; Pollak 2016; Wadd et al. 2024).

Dilemmas related to older adults’ alcohol use within care and nursing homes arose when respect for residents’ autonomy conflicted with health and safety or the comfort of the resident or others around them, in which case the other residents were usually prioritized. This decision was more complicated within wet care homes that acknowledged that residents were limited in where they were able to live. Within RCSs that housed comparatively older adults compared to the other included studies (Emiliussen et al. 2021; Klein and Jess 2002), heightened concerns around alcohol inducing falls and interacting with medication led to emphasis on rules, such as allowing alcohol use only on special occasions or based on doctors’ orders. Issues cited by staff and occasionally managers around managing alcohol use in RCSs were mainly only present when residents lacked mental capacity to control their own intake (Emiliussen et al. 2021; Wadd et al. 2024) or when alcohol use was problematic in that it had an obvious effect on individuals or others around them (Johannessen et al. 2021). For example, leaving alcohol unattended presented problems particularly for individuals with alcohol dependency or brain injury (Wadd et al. 2024).

Within the rare situations of issues related to alcohol use of residents, wet RCSs (McCann et al. 2017; Nixon and Burns 2022) generally demonstrated adaptive, appropriate and measured solutions to manage alcohol use such as the creation of individualized care plans (Nixon and Burns 2022), facilitated by specialist skills and experience (McCann et al. 2017). Whilst staff within both studies did not deny challenges, including antisocial behaviour and residents seeking more alcohol than was agreed upon and provided to them, it was clear that residents recognized the RCS as somewhere they were accepted and supported (McCann et al. 2017; Nixon and Burns 2022). When limitations were enforced publicly and paternalistically, residents felt frustrated and infantilized (McCann et al. 2017). However, when gentle limitations were enforced when necessary with tact, communication, trust and respect for dignity, residents were more accepting (McCann et al. 2017), eliciting positive outcomes including decreased use of health and social care and the criminal justice system (McCann et al. 2017). Perhaps related to the wide age range of residents in the wet care home, some welcomed the support of staff intervention to restrict their alcohol use, whilst others resented the lack of autonomy (McCann et al. 2017).

On the contrary, ‘regular’ RCSs were less equipped to deal with such challenging situations and dilemmas, and, in the absence of specialist training or guidance, decision-making was informally attributed to front-line staff (Emiliussen et al. 2021) who were trusted by their managers to make decisions (de Graaf et al. 2023b). Only one study described a protocol for dealing with problematic alcohol use in the RCS, although it was unclear how formal this was (Pollak 2016). Furthermore, stricter policy within RCSs meant that residents who were physically unable to purchase alcohol themselves relied on staff (Johannessen et al. 2021) or family (de Graaf et al. 2023b; Emiliussen et al. 2021; Johannessen et al. 2021) to supply them with alcohol, providing a legal grey area, particularly if alcohol was permissible within residents’ rooms (Emiliussen et al. 2021). Occasionally a desire for alcohol was projected onto the older person without an indication of their desire for it, particularly in residents who lacked capacity to communicate their choice. Staff sometimes followed specific principles to guide their decision-making, including the ‘dignity of risk’ principle, namely, balancing the potential risks and benefits of alcohol consumption by allowing the dignity to take risks and reducing risk in other ways, and the principle of ‘least restrictive option’ (Wadd et al. 2024). Contradictorily to a respect for residents’ autonomy and choice, efforts to manage alcohol intake by staff were most often enforcements on the environment such as limiting alcohol consumption, and the only psychological approaches identified were the co-creation of care plans (Nixon and Burns 2022) and the availability of Alcoholics Anonymous meetings on site (Chambers 2020; Pollak 2016). There were some cases of deceit where staff swapped alcoholic beverages for non-alcoholic options without the knowledge of a resident with alcohol dependency or dementia (de Graaf et al. 2023b; Pollak 2016). The benefit of informal decision-making was that decisions tended to be specific to the situation and the individual, although informal decisions could also be subjective and inconsistent across staff.

Most existing communication around residents’ alcohol use was between care staff and family members and did not always involve the older adult (de Graaf et al. 2023b). Where assumptions were made about a resident’s need without their involvement, trust in staff could be impaired. Current and historic alcohol consumption information was often not collected on admission (Klein and Jess 2002; Wadd et al. 2024), leading to a lack of knowledge of staff (de Graaf et al. 2023b; Klein and Jess 2002). Also, given that residents were not always honest about their consumption, perception of a resident’s alcohol intake could differ drastically between stakeholders (Chambers 2020), and staff and family members sometimes employed covert methods of accessing residents’ alcohol consumption such as checking financial records (Chambers 2020).

The attitudes and subsequent facilitation of staff and family members were inherently influenced by their own experiences, attitudes, beliefs and drinking history and habits (Johannessen et al. 2021; Wadd et al. 2024). Consequently, stakeholders were not always in agreement about the decisions made about a resident’s alcohol intake. Often, family members were the largest advocates of residents’ alcohol use and limited staff ability to reduce the residents’ alcohol consumption by supplying residents with alcohol without the knowledge of staff (Emiliussen et al. 2021; Johannessen et al. 2021; Pollak 2016). Meanwhile, staff and managers more strongly weighed the needs of individual residents with communal living (Emiliussen et al. 2021), risks (McCann et al. 2017; Nixon and Burns 2022; Wadd et al. 2024) and liability concerns (Wadd et al. 2024).

Discussion

This systematic review aimed to examine the perspectives of relevant stakeholders on alcohol use in RCSs by older adults. Whilst the motivations, influencing factors and social acceptability relating to alcohol use of older adults are relatively consistent across various RCSs, the need to balance individual rights with risks and the importance of guidance for decision-making became increasingly relevant within care and nursing homes with the decreasing psychological and physical capacity of residents. In such situations, there is a lack of guidance and training to support staff in responding. Older adults’ alcohol use is generally not problematic (Bareham et al. 2020) and older adults regulate their own intake with limitations and rules (Bareham et al. 2019; Kelly et al. 2018), although residents in RCSs are resourceful in fulfilling their desired alcohol consumption despite prohibitory policy and constraints. When life is otherwise potentially very different, alcohol use offers a continuation of existing routines and rituals (Bareham et al. 2019; Kelly et al. 2018). Choice was inherently linked to a feeling of being at home and belonging, and decisions around alcohol use offer residents a choice in a setting where opportunity for choice is scarce (Bareham et al. 2020). Furthermore, RCSs are situated within a context of a broader societal acceptability of alcohol use (unless it affects others [Bareham et al. 2019; Muhlack et al. 2018]) compared to other recreational drugs. Also, alcohol often enriches the lives of older adults, facilitating socializing (Bareham et al. 2019; Bareham et al. 2020; Kelly et al. 2018; Parke et al. 2018), quality of life (Bareham et al. 2020; Kelly et al. 2018) and enjoyment (Bareham et al. 2019; Kelly et al. 2018). Thus, approaches to alcohol use within RCSs should balance consideration of the physical health risks with consideration of the positive social and psychological impacts experienced by many residents, as consequently many older adults and their family and carers are motivated to encourage alcohol use in RCSs. Indeed, prohibitory policies placed more of the decision-making burden on staff and family instead of on the older adults themselves, reflecting a wider debate around autonomy and choice within RCSs in decisions around older adults’ care (Hedman et al. 2019). However, for a minority of older adults, alcohol use creates negative outcomes that impact staff and residents within RCSs, and overall there is a lack of reflection amongst older adults on their own intake (Bareham et al. 2019; Bareham et al. 2020; Muhlack et al. 2018). Thus, person-centred decision-making around older adults’ alcohol use is needed, facilitated by rapport with residents (Bareham et al. 2020) and adequate training for staff in dealing with alcohol misuse (Bareham et al. 2020).

Unique challenges associated with RCSs included ensuring communication with and involvement of all relevant stakeholders in decision-making around residents’ alcohol use, inherent limitations around control and autonomy, and managing differing levels of alcohol use amongst residents. Reports of deceit being practised towards individuals with dementia or alcohol dependency raise the ethical question of whether deceit is ever appropriate in health and social care. Depending on the ethical theoretical underpinning adopted, being deceitful towards residents with dementia may be morally acceptable when prioritizing the best interests and care of the resident as opposed to the convenience of lying (Cantone et al. 2019). Such challenges were not present within independent living settings, which may instead be comparable to private residences, although this difference could be owing to differences in study aims. Thus, research is needed in independent living RCSs that includes wider stakeholders such as wardens and family members, to explore their involvement in residents’ alcohol use.

The current review identified a lack of guidance for staff navigating residents’ alcohol use, resulting in subjective decision-making that could promote unfairness across residents and even inequalities by SES, given that facilitation and acceptability could be influenced by SES. In the absence of national guidance (Wadd et al. 2024), RCSs often create their own policies, which vary greatly from each other and result in a wide variation in care. Furthermore, a lack of national guidance leaves front-line staff to balance policy for person-centred care, which means including residents in decisions about their care (Care Quality Commission 2014a), safe care and treatment, which means avoiding preventable harm or risk of harm (Care Quality Commission 2014b), and the Mental Capacity Act, which ensures that staff act in the best interests of residents only when they lack capacity to make decisions (Department of Health 2005). In practice, this means balancing legislation that encourages residents to make their own decisions even when they are unwise for their health with legislation that is risk averse and assigns responsibility to staff for residents’ wellbeing. Existing guidance for care staff provides advice on navigating alcohol use with residents, including those with alcohol dependency (Rota-Bartelink 2011a, 2011b; Wadd et al. 2023), communication limitations (Wadd et al. 2023) and limited capacity (Wadd et al. 2023), although it does not incorporate the input of wider stakeholders including family members, who this review found are often the most influential in determining alcohol use of residents.

The literature on wet care homes (McCann et al. 2017; Nixon and Burns 2022) demonstrated the utility and success of flexible guidance and principles including harm reduction, shared decision-making and person-centred care. However, this literature also highlighted the specialist skills and knowledge needed to care appropriately for individuals with alcohol dependency (McCann et al. 2017). Thus, through adequate training and protocols, RCSs should be better prepared to care appropriately for individuals with alcohol dependency or at least be capable of referring residents to a RCS that is, to ensure inclusivity of all residents regardless of their drinking history. Standardized guidance around managing alcohol use in older adults in RCSs on admission would help to promote fairness and equality across all residents, promote involvement of all stakeholders and support staff in making challenging decisions. For example, a harm reduction intervention for residents with alcohol dependency and associated antisocial behaviour that provided specialist and highly personalized structured activities and behavioural management was found to be cost-effective in reducing depression, anxiety, problematic drinking and number of alcoholic drinks consumed (Rota-Bartelink 2011a, 2011b). Whilst most of the subjective decision-making observed within the current review already appeared to be person-centred, it was often influenced by the attitudes and experiences of the decision-maker and did not sufficiently involve all relevant stakeholders, as is required for successful person-centred care (de Graaf et al. 2023a). Also, the current review identified a lack of psychological intervention and reflection of residents on the purpose and risks of their own alcohol use. Thus, shared decision-making could provide an opportunity for a brief intervention (Gordon et al. 2003) to encourage reflection on when alcohol meaningfully adds to residents’ experience and quality of life balanced against the health risks, and when a non-alcoholic alternative would be just as meaningful, which is also not accounted for in the existing guidance (Rota-Bartelink 2011a, 2011b; Wadd et al. 2023). For example, a shared decision to offer alcohol only at mealtimes could utilize the positive social effects of drinking alcohol identified within this review whilst also attenuating for some of the increased risk of mortality caused by alcohol use (Ortolá et al. 2024).

The main strength of the current review was the comprehensive search strategy, including grey literature, which accessed perceptions of a range of stakeholders and included residents across all drinking histories, including dependency. Compared to previous systematic reviews’ justification for excluding studies of residents with alcohol dependence as they are ‘strongly encouraged to abstain from drinking’ (Bareham et al. 2019; Bareham et al. 2020), the current review acknowledged harm reduction as a more pragmatic, safe and inclusive approach to managing alcohol use. The main limitations of the current review were that no translation services were available, which may have affected the ability to include all eligible studies. Also, a lack of open data meant that the only quotations available for extraction were those selected by the relevant studies’ authors to be presented within their manuscripts; thus, it is uncertain whether the conclusions made by the current review reflect the constituent datasets of the included studies. An increase in data sharing for qualitative research will facilitate reviewers of qualitative studies in comprehensive synthesis.

In conclusion, it is pragmatic to accept that older adults in RCSs should be involved in shared decision-making with other relevant stakeholders, including staff and family, to determine their own alcohol use. Facilitation of alcohol in RCSs respects the choice and autonomy of older residents, whilst adopting a harm reduction approach helps to minimize the risks towards residents and those around them. As the available literature within sheltered accommodation and assisted living facilities mainly focused on residents’ perspectives only, themes within these included studies focused on residents’ motivations and influences relating to their alcohol use. Thus, further research is needed to include the perspectives of all relevant stakeholders and concerning independent living in RCSs, specifically around the intersectionality between SES and how different policies across RCSs relate to means, opportunity and capacity to consume alcohol. Future research on independent living in RCSs should also focus on the perspectives of family members, in the absence of continual formal care provision. Also, more research is needed within wet care homes to explore the experiences of commissioners, managers and staff around managing alcohol use in these challenging settings, which balance caring for residents of a wide range of ages. Crucially, practical guidance is needed on managing alcohol use in older adults that includes all stakeholders and provides the opportunity for psychological intervention.