Health and social care providers' perspectives of older people's drinking: A systematic review and thematic synthesis of qualitative studies.

BACKGROUND: alcohol may increase risks to late-life health, due to its impact on conditions or medication. Older adults must weigh up the potential risks of drinking against perceived benefits associated with positive roles of alcohol in their social lives. Health and social care workers are in a key position to support older people's decisions about their alcohol use.
OBJECTIVE: to systematically review and synthesise qualitative studies exploring health and social care providers' views and experiences of older people's drinking and its management in care services.
METHOD: a pre-specified search strategy was applied to five electronic databases from inception to June 2018. Grey literature, relevant journals, references and citations of included articles were searched. Two independent reviewers sifted and quality-appraised articles. Included study findings were analysed through thematic synthesis.
RESULTS: 18 unique studies were included. Four themes explained findings: uncertainty about drinking as a legitimate concern in care provision for older people; the impact of preconceptions on work with older adults; sensitivity surrounding alcohol use in later life; and negotiating responsibility for older adults' alcohol use. Discipline- and country-specific patterns are highlighted.
CONCLUSIONS: reservations about addressing alcohol could mean that service providers do not intervene with older adults. Judgements of whether older care recipients' drinking warrants intervention are complex. Providers will need support and training to recognise and provide appropriate intervention for drinking amongst older care recipients.

A range of complex factors influence care providers’ perceptions of whether addressing alcohol should be part of their practice.

Preconceptions of at-risk drinking groups and older adults’ capacity to change shape the approaches of service providers.

Discussion of alcohol is avoided because it is viewed as a sensitive topic.

Care providers’ perceptions of their own roles and competing priorities influence how they address older people’s drinking.

Some care providers will need support and training to provide appropriate intervention for older adults’ drinking.

Introduction

Alcohol use is a leading modifiable risk factor for disease [1]. Levels of use defined as lower risk for the general population within alcohol use guidelines may become hazardous or harmful in later life [2, 3]; where use could lead or has led to physical, psychological or social harm. This is because tolerance for alcohol decreases with age [4-7], and many older adults live with medical conditions, and may take medications to manage them; either of which can be affected by drinking [8]. Within high-income countries, up to 87% of older people use alcohol, with up to 45% at risk of health complications resulting from their intake [9]. This large at-risk group causes greater strain to healthcare systems than the smaller group diagnosed with alcohol dependence [10]. As the population ages, these systems will face increasing pressure from the consequences of older people’s drinking [9].

Alcohol can play a central role in older people’s social lives, positively contributing towards their quality of life [11]. There have also been suggestions of health benefits at lower levels of intake [12-18]. Such findings likely reflect characteristics of non-drinking groups, who have often ceased drinking due to pre-existing health complaints that mean they are more likely to experience health difficulties [19]. Nonetheless, suggested health benefits may motivate older adults’ alcohol use [11]. Older people must weigh up potential risks against the benefits they perceive from drinking in making decisions regarding their intake. However, most older adults who experience health consequences from their drinking do not recognise alcohol’s role [2]. Awareness of what constitutes lower risk alcohol use is poor amongst the older age group [20].

Health and social care settings provide a context to screen for hazardous use, identify risks associated with medicine use or health state, and where older people can be supported to make healthier decisions regarding their drinking [21]. Older adults are responsive to interventions to address alcohol use [22]. However, care providers often fail to identify older adults’ hazardous drinking or do not deliver appropriate intervention [20]. It has been suggested that practitioners’ attitudes may affect this area of preventive care [6].

This review synthesises qualitative studies reporting formal health and social care providers’ views and experiences of older people’s drinking within high-income countries. An understanding of providers’ perceptions is essential to recognise how they can be supported in alcohol-related practice with older adults. The aim of this review is to identify and understand issues that may influence care providers’ efforts to address hazardous or harmful drinking amongst older care recipients.

Methods

Detailed methods have been published elsewhere [11].

Search strategy

Five bibliographic databases were searched from inception to June 2018 (OVID: Medline (1946), PsychINFO (1806), Scopus (1960), EBSCO Cumulative Index to Nursing and Allied Health Literature (CINAHL, 1984), ProQuest Applied Social Sciences Index and Abstracts (ASSIA, 1987)). The subject of interest was defined in terms of study population, phenomena of interest and context of study, as recommended by the Joanna Briggs Institute [23]. Search terms and eligibility criteria were developed in accordance with this definition. Database-specific headings and key words were developed relating to the concepts “older adults”, “drinking”, “qualitative” and “perceptions and experiences”. Keywords were mapped to subject headings for each database, which were exploded, focussed and combined appropriately. This produced a search strategy optimised for sensitivity (tested for inclusion of known relevant articles) and specificity (minimising identified articles irrelevant to the topic). The following grey literature sources were searched, applying key terms: NHS evidence, Open Grey and Dissertation Abstracts International. The full search strategy applied to each database within this review is available through our Prospero registration [24]. References and citations of included articles were searched for further eligible articles.

Eligibility criteria

Published studies presenting qualitative analysis in any language were included, with the exception of reviews and case studies. No time limits were applied.

Included studies focussed on:

formal health and social care workers’ perceptions and experiences of older people’s drinking, defined as aged 50 years and over.

views of individuals living in OECD high-income countries. Alcohol use is prevalent and problematic amongst these older age populations [9], and countries’ care systems have comparable resources to support health promotion [25].

Studies were excluded if:

they focussed on individuals known to be dependent on alcohol, as treatment populations are strongly encouraged to abstain from drinking.

alcohol use could not be distinguished from other substance use.

Data extraction and quality assessment

Following electronic de-duplication, two independent reviewers screened papers for relevance based on titles and abstracts, then full text papers were assessed for inclusion. Discrepancies were discussed and resolved. Non-English full text papers were translated into English by native speakers. Details of the study setting, participants, methods and study data were extracted from included articles. Study quality was assessed by two independent reviewers using Saini and Shlonsky’s Qualitative Research Quality Checklist [26]. This tool guides evaluation of methodological rigour and appropriateness, and level of detail and interpretation in presented results. Studies were not excluded on the basis of quality appraisal, as poor reporting is not necessarily indicative of badly conducted research [26]. However, assessing quality prevents unreliable results from influencing review findings [27]. Key limitations and comments on richness of presented findings are detailed for each study in Table 1 (see Supplementary Table S1 for full appraisal) and summarised within our study descriptions to give a sense of limitations and richness of available data.

Table 1

Brief descriptive summaries of included studies with key limitations and comment on richness identified in quality appraisal

Article and country	Aims	Sample	Data collection methods and analysis	Author-identified key themes	Key limitations and comment on richness from quality appraisal
Andersson and Bommelin [32], Sweden	To examine domiciliary care managers’ understanding of older people’s hazardous use of alcohol, in the context of aid assessment.	n = 5, occupation = domiciliary care managers.	Semi-structured interviews, analytical systematic text condensation	Perilous use or misuse; neglected social needs; assistance based on the elders’ own request or initiative	Small sample size, which is not reasoned in terms of data saturation; thin description of findings, lacking contextual detail; reporting of findings was predominantly descriptive.
Andersson and Johansson [33], Sweden	To describe how municipal elderly care providers perceive, manage and treat older people with alcohol use and abuse.	n = 6, occupation = elderly residential care providers and domiciliary care providers.	Semi-structured interviews, thematic analysis	Alcohol policy; an individual’s own choice; medicine and alcohol; the staff’s view of the importance of alcohol to the elderly; the care recipients’ background; relatives and over reporting; the future	Inconsistent transparency in reporting; small sample size, which is not reasoned in terms of data saturation; unclear to what extent findings were grounded in the experiences of participants; rich description of findings with thick contextual detail.
Broyles et al. [43], United States of America	To identify the potential barriers and facilitators associated with nurse-delivered alcohol screening, brief intervention and referral to treatment for hospitalised patients.	n = 33, occupation = medical-surgical nurses.	Semi-structured focus groups, constant comparison	Anticipated barriers; suggested facilitators	Researcher biases not discussed in reporting; thinner description of findings - limited contextual detail provided to support understanding of reported findings.
Claiborne et al. [39], United States of America	To identify the primary care practice patterns relevant to patients’ alcohol problems; to identify barriers and incentives for use of particular Veteran’s acute care guidelines for screening and referral for evaluation and treatment of these problems.	n = 31, occupation = medical physicians, medical nurse practitioners, registered nurses or nurse practitioners, behavioural health providers, administrative staff and support staff.	Structured one-to-one interviews, constant comparison	AUDIT-C screening process; identifying alcohol problems; referral for further evaluation and treatment; follow-up with patients; perception of behavioural health provider	Inconsistent transparency in reporting; researcher biases not discussed in reporting; thin description of study findings – some contextual details supplied, but just one supporting quote and findings quantified in places.
Dare et al. [42], Australia	To investigate how issues related to role legitimacy, role adequacy and role support acted as barriers and/or enablers to community pharmacists’ practice in delivering health information and advice about alcohol to older clients.	n = 14, occupation = community pharmacists.	Focus groups, inductive thematic analysis	Professional activities; professional attributes; pharmacist/client interactions and relationships; infrastructure, support and materials	Few data items compared with the amount that might be expected for a focus group study; unclear how the applied theoretical framework was incorporated within the inductive approach to analysis; thick description of findings, exploring negative cases and trends.
Darwish and Fyrpihl [34], Sweden	To investigate how care workers handle and interpret alcohol problems in elderly service users.	n = 9, occupation = care managers, nurses and auxiliary nurses.	Semi-structured interviews, thematic analysis	The interpretation of an alcohol problem; the ethical dilemma; flaws within the field; strategies and policies	Sample size is not justified in terms of data saturation; researcher biases not discussed in reporting; quotes provided do not consistently support authors’ narrative; findings appear to be imposed by the theoretical framework with little support from participants’ perspectives; thinner description of findings, with few contextual details presented.
Gunnarsson [35], Sweden	To conduct an exploratory study of the perspectives of domiciliary care providers working with the elderly with substance abuse problems.	n = 11, occupation = domiciliary care providers.	One-to-one interviews, thematic analysis	Assessment of needs and substance abuse; the domiciliary carer’s every day	Inconsistent transparency in reporting; researcher biases not discussed in reporting; thick descriptions of study findings.
Gunnarsson and Karlsson [36], Sweden	To explore domiciliary care assistants’ perceptions of drinking in later life (not explicitly reported).	n = 34, occupation = domiciliary care manager, domiciliary care nurses, auxiliary nurses and nursing assistants.	Focus groups and one interview, thematic analysis	The care takers’ opinion of the work with elderly and alcohol problems: How the care takers act; care planners’ view of elderly and alcohol; the care staffs’ talking about their work with the elderly with alcohol problems	Inconsistent transparency in reporting; researcher biases not discussed in reporting; thick descriptions of study findings.
Herring and Thom [30], [31], United Kingdom (England)	a) To explore policy and practice regarding the purchase of alcohol for older clients of domiciliary carers in three local authorities in the Greater London area. b) To assess the current and potential role of domiciliary carers in the identification and response to problems associated with alcohol use and misuse in older people.	n = not reported, occupation = domiciliary carers and their managers.	Semi-structured interviews, focus groups and written responses to postal questionnaires. a) qualitative analysis b) grounded theory approach	a) (Findings organised under pseudonyms of local authorities studied) b) Alcohol policy; domiciliary carers’ perceptions of alcohol misuse; domiciliary carers’ ideas about why older people may misuse alcohol; spotting alcohol misuse: what do domiciliary carers think are the signs?; response to alcohol misuse; “Like a daughter”; The relationship between domiciliary carers and their clients	a) Reporting was not transparent; researcher biases not discussed in reporting; little detail provided on the study sample; some contextual details presented in reporting trends, providing a thicker description of findings. b) Reporting was not transparent; researcher biases not discussed in reporting; thin description of finding.
Johannessen et al. [47], Norway	To investigate health personnel’s perceptions and experiences of alcohol and psychotropic drug use among older people and to what extent this is an issue when services are planned for and implemented	n = 16, occupation = district nurses, occupational therapists and physiotherapists.	Semi-structured interviews, qualitative content analysis	State of practice; a desire to improve services	Researcher biases not discussed in reporting; thin description of findings.
Johannessen et al. [40], Norway	To investigate general practitioners’ experiences and reflections on use and misuse of alcohol and psychotropic drugs among older people, and to what extent this is an issue in treatment.	n = 11, occupation = general practitioners.	One-to-one interviews, phenomenological-hermeneutical method	The GP’s opinion of older people’s alcohol and psychotropic use; the GP’s practice	Researcher biases not discussed in reporting; thin description of findings.
Johannessen et al. [41], Norway	To explore how health professionals experience their participation in a study in which they collected data on alcohol and psychotropic drug use among patients treated in old-age psychiatry departments, and subsequently how they experienced their work day after the study ended	n = 15, occupation = nurses, psychiatrists, social workers.	Focus-groups and individual interviews; manifest qualitative content analysis	Experiences with participation; consequences of participation	Sampling criteria were not discussed; no reflexivity evident in reporting, and it was unclear which elements of the study findings were grounded in the data versus in researchers’ ideas and assumptions; thin description of findings.
Koivula et al. [45], Finland	To examine how the alcohol use of elderly domiciliary care clients affects the daily work of domiciliary care professionals and how the professionals act to support the drinking client	n = 10, occupation = domiciliary care professionals.	Semi-structured interviews, method of analysis not reported	Supporting life management of the client; the lack of qualifications in tackling clients’ drinking; the need for multi-professional collaboration	No details provided regarding study recruitment, sampling or methods of analysis; narrative appeared to be influenced by researchers’ understanding of pre-existing literature, however lack of detail of study methods meant it was difficult to gauge to what extent this was the case; thin description of findings.
Millard and McAuley [46], United Kingdom (Scotland)	To explore: (1) how clients’ alcohol problems were identified (2) was it the domiciliary care providers work role to raise a possible alcohol problem with a client (3) whether domiciliary care providers had sought help for a client with alcohol problems, and if there were any barriers (4) were there any gaps in services for older people with alcohol problems, and if so, how might they be filled?	n = 90, occupation = domiciliary care staff and domiciliary care managers	Focus groups, method of analysis not reported	None reported. To summarise findings: Trusting relationship between domiciliary care workers and clients; domiciliary care workers’ perceptions regarding the client’s alcohol consumption; barriers to involvement in day care or residential settings secondary to the client’s alcohol usage; the impact of Scottish culture	Reporting was not transparent; researcher biases not discussed in reporting; thin description of findings.
Serbic and Sundbring [37], Sweden	To investigate how residents’ alcohol consumption is treated and handled by nursing staff.	n = 6, occupation = trained nurses working in elderly care and elderly care workers.	Semi-structured interviews, thematic analysis	A picture of consumption; self-determination and quality of life; cumbersome situations; different experiences	The sample size is not justified in terms of data saturation; researcher biases not discussed in reporting; thick description of findings.
Severin and Keller [38], Sweden	To explore the domiciliary care staff’s experiences of working with older patients who have alcohol problems.	n = 6, occupation = domiciliary care nursing assistants.	Semi-structured interviews, inductive thematic analysis	Self-determination as an obstacle – dilemmas at work; adaptation and flexibility as a means of management strategy; support that is lacking in the work—support in the current situation	Inconsistent transparency in reporting; quotes provided were not always supportive of the authors’ narrative; thick description of findings.
Shaw and Palattiyil [44], United Kingdom (Scotland)	To explore social work practitioners’ awareness of alcohol misuse in older people, and their attitudes towards the current support services.	n = 18, occupation = social work practitioners.	Semi-structured interviews, thematic analysis	Extent of the problem; difficulties identifying the problem; reasons for alcohol problems among older people; unmet need among older people with alcohol problems; more effective service provision	Inconsistent transparency in reporting; researcher biases not discussed in reporting; thin description of findings.

Data synthesis

A thematic synthesis of included studies was conducted (a process diagram is available elsewhere [11]). Methods for synthesis were based on Braun and Clarke’s principles of thematic analysis [28]. The review team familiarised themselves with findings of each study during full text screening. During this phase, the lead author listed ideas and potential codes from primary study findings. Compiled codes were comparable to second- and third-order constructs described in meta-ethnography. Second-order constructs are interpretations and themes derived from primary data, specified by authors of included studies. Third-order constructs are interpretations identified by the review team that further explain findings within and across primary studies [29]. Recurring codes, explaining findings across studies, were developed into a candidate framework of themes that explained issues affecting care providers’ efforts to address hazardous or harmful drinking amongst older care recipients. NVivo (version 11) was used for data management. The lead reviewer recorded analytical notes during this process, detailing explanations and patterns within each theme. The thematic framework was further refined to ensure that it reflected views and experiences conveyed across included studies, and defined to form the theme descriptions presented as our findings. Excerpts from included studies (supporting quotes from study participants and extracts from study authors’ narratives) were identified to present as examples. Developing themes were discussed amongst the research team to inform data interpretations.

Results

Literature search and study descriptions

Eighteen papers met eligibility criteria (see Figure 1), reporting 17 unique studies ([30, 31] were analyses of the same data set). Brief descriptive summaries of included studies are reported in Table 1, with additional detail supplied in Supplementary Table S1. Seven articles were theses translated for inclusion [32-38] (indicated in Supplementary Table S1).

Figure 1

PRISMA flow diagram depicting the flow and number of studies identified and then excluded at each stage during identification of papers for inclusion in this review

The synthesis included data from 329 care providers. Job roles included family physician/general practitioner [39, 40], psychiatrist [41], community pharmacist [42], general practice nurse [39], district nurse [34, 36], specialist nurse [35, 37, 39, 41, 43], social worker [41, 44], domiciliary carer [30, 33, 35, 37, 45, 46], residential home carer [33, 35, 36, 46], domiciliary care manager [30, 32, 34–36, 46], healthcare assistant [34–36, 38], physiotherapist [47], occupational therapist [47] and behavioural health provider [39]. Twelve studies explored the perspectives of multiple health and social care workers with different roles [30, 33–37, 39, 41, 45–47].

Care providers worked with older people in their homes [30, 32–38, 44–47], residential care [32–37, 46], general practice [39, 40], medical surgical ward [43], old age psychiatry unit [41] and community pharmacy [42].

Where stated, practitioners’ age ranged from 18 to 70 years, gender was mostly female (average 88% of sample) and reported ethnicity was majority white. Years in practice ranged from 0 to 36.

Fourteen studies detailed the age group discussed by providers. Most discussed people aged 65 years and over [32, 33, 35, 36, 38, 40, 41, 43, 44, 46, 47]. Two studies discussed people aged 60 and over [42, 45], and one study discussed those aged 50 and over [39]. Different studies explored care providers’ understanding of older people’s use and misuse of alcohol [31–40, 44–47], perceived roles and approach to older people’s alcohol use [30-47], and factors affecting their work [30–36, 38–47]. All studies referred to a range of drinking practices, including drinking at any level [30, 33, 35, 37, 39, 41, 42, 45–47] and misuse of alcohol [31–35, 38–41, 43, 44, 47]. Four studies included some discussion of dependent drinkers [32, 39, 43, 45]. Material addressing each topic and level of use was developed from data from a range of providers working with older adults in primary care, specialist care and in older people’s homes (although material regarding understanding of alcohol use did not incorporate perspectives of providers working in specialist care). Further details of contributing material are presented in Supplementary Table S2 for level of use, and Supplementary Table S3 for topic.

Studies were conducted in the UK [30, 44, 46], the USA [39, 43], Sweden [32-38], Norway [40, 41, 47], Finland [45] and Australia [42].

Where reported, samples were recruited either purposively [32, 33, 36, 41, 47] or opportunistically [30, 34, 37–40, 42]. Data were collected through in-depth/semi-structured interviews and focus groups. A range of approaches were applied in analyses. These included thematic analysis, constant comparison and grounded theory.

The main quality limitations related to small samples (by qualitative standards) not determined by data saturation; and lack of transparency in reporting, particularly limited contextual detail and depth.

Themes

Four themes were formulated by review authors (representing third order constructs) that explained issues affecting care providers’ work to address older care recipients’ drinking, conveyed across included studies. Supporting excerpts are presented in Table 2.

Table 2

Supporting quotes for presented themes

Theme	Supporting quotes
(1) Uncertainty about legitimacy of drinking as a concern in care provision for older people	(i) Care providers came to recognise the rising prevalence of at-risk drinking amongst the older age group when they took an explicit focus upon addressing older adults’ drinking in their work. This was the case amongst care providers in an old age psychiatry department, who had come to focus upon addressing care recipients’ drinking having recently been involved in a study where they assessed alcohol use:
	“After participating in the data collection for the previous project, the informants expressed in the group discussions and in the interviews that they had become more aware of the importance of the topic and that they wanted to maintain a special focus on elevated alcohol and psychotropic drug use in the treatment of patients referred to old-age psychiatric departments […] The informants also reported that they had learned a lot about elevated alcohol and psychotropic drug use through their participation in the project” (Comment on care providers working in old age psychiatric departments, Johannessen et al. [41])
	(ii) Moderate alcohol use could be seen to contribute towards older people’s quality of life, and was therefore accepted by some care providers:
	“Several of the interviewees explain that […] they realise that their [clients”] quality of life is increased when they […] get to drink “their little whiskey before bedtime”. Hence, they don’t restrict somebody’s consumption as often as they would like to, as long as the individual doesn’t harm themselves, or anyone else.’ (Comment on elderly care nurses and carers, Serbic and Sundbring [37])
	(iii) Care providers discussed the roles of alcohol in coping with loss of purpose, which they associated with retirement, bereavement and loneliness in later life:
	“The informants experienced that many of their older patients were lonely, and therefore, used alcohol […] to reduce their strain. Structural changes in their lives and in society, such as children having moved out or were too busy with their own lives, loss of friends, dependency because of poor health, and few meeting places for older people, were seen as reasons for loneliness.” (Comment on GPs, Johannessen et al. [40])
	(iv) Care providers discussed how alcohol’s role in coping could create a pathway to alcohol dependence in later life:
	“These service users begin using alcohol to cope with the feelings of emptiness and grief following their partner’s death and continued to drink until it has become a way of life and they are effectively dependent upon alcohol. They also have very little else to fill their day.” (Social work practitioner, Shaw and Palattiyil [44])
	(v) When care providers perceived that an older person’s alcohol misuse warranted intervention, this was almost always as a result of emerging consequences or indications of alcohol dependence rather than any attempt at harm prevention:
	“Home carers had a very ‘black and white’ view of alcohol-related problems: a person was either alcoholic or did not have a problem. There seemed to be little understanding that some older people may experience alcohol-related problems when drinking a moderate amount, for example, because they have impaired balance.” (Comment on domiciliary care providers, Herring and Thom [31])
	(vi) and (vii) Older clients’ intoxication presented particular challenges to domiciliary carers, where older people could consume alcohol in their work setting:
	“A concern raised by home carers was the risk of fires and accidents when older people smoked and drank. In one case, they were first alerted to the possibility that the client was drinking when: ‘She started to drop her cigarettes badly … she didn’t realise and she actually burnt to death.’” (Comment on domiciliary carers, Herring and Thom [31])
	“Many of the staff are harmed as well. Everyone can’t deal with it. Some get scared. I haven’t experienced any fear, but some of my colleagues do get scared to an extent where they don’t want to go to work.” (Nurse, Darwish and Fyrpihl [34])
(2) The impact of preconceptions on work with older drinkers	(i) Care providers’ expectations regarding which patients or clients were more likely to misuse alcohol and how this might manifest could present a barrier to identifying problematic use:
	“Another general expectation is that elderly don’t consume alcohol, especially not older women, according to the home care managers. One home care manager describes one elderly woman who she first thought had a problem with her memory and the relatives responded with: ‘Memory problems! She was drunk!’ Intoxication with elderly women is often accepted with confusion, if ‘the smell of alcohol’ isn’t there or there aren’t bottles on the table, or empty bottles found around the house.” (Comment based on domiciliary care managers, Gunnarsson and Karlsson [36])
	(ii) Care providers saw older people’s drinking practices to be irreparably ingrained by later life:
	“[nurse 1] Our population is probably mid-50s to older- it’s something they’ve been doing for 25–30 years… at that point they don’t think they have a problem, it’s just normal to them.” “[nurse 2] Or it’s already too long. They’ve already got the problems that go with it [alcohol use], and think, why bother?” (Medical surgical nurses, Broyles et al. [43])
(3) Sensitivity surrounding alcohol use in later life	(i) Care providers saw that rapport built with the client through interaction during care provision facilitated successful discussion of alcohol misuse:
	“We understand that when the care takers have built a good relationship with the care recipient, they can be deemed as significant by the care recipient, and the care takers can then approach them with questions regarding alcohol. We interpret this as the care takers trying to show empathy and create relationships on personal levels in order to help the elder. And thus they create trust with the elder enabling this topic to be discussed.” (Comment on care managers/nurses, Darwish and Fyrpihl [34])
	(ii) Care providers felt that rapport built with their older patients and clients was essential to other care tasks, and could be damaged where alcohol-related discussion was unsuccessful and led to the older person becoming offended. This could lead to the older person becoming resistant to their care provider and refusing to accept care:
	“You try to go and say “Let’s discuss this [alcohol misuse],” and they will swear at you and throw stuff at you, and tell you to “get out”… And the next time you come back to take care of a medical thing, they’re spitting at you, because they said, “I told you I don’t ever want to see you again.” (Medical-surgical nurse, Broyles et al. [43])
	(iii) Care providers reported how sensitivity regarding alcohol misuse can be accentuated amongst social groups where drinking is seen to be less socially acceptable:
	‘It isn't always that these women are aware that they are alcoholics. “You don't get it out of them that they have an abuse problem, but no I just drunk some beer every now and then… there is no absolute… it's more hush hush. Even amongst their relatives. That mum. . . dad can drink, but not necessarily mum in the same way, that's harder to accept. Those who do have relatives, it isn't everyone who does have them”’ (Comment on domiciliary care providers, Gunnarsson [35])
	(iv) and (v) Care providers explained that older care recipients’ sensitivity represented a barrier to discussing or recording increasing risk drinking in care settings:
	“Usually we do not write it [hazardous/harmful drinking] in the plan, because it’s stigmatising. We have no right to stigmatise anybody. For the moment we have one person who has not used alcohol for many, many years. And still when this client visits the hospital, doctors write, every time, “heavy user of alcohol”, just because it’s mentioned in some document. And so it keeps going. I myself am really cautious, I do not want to stigmatise anybody.” (Domiciliary care provider, Koivula et al. [45])
	“No, no direct questions [...] Nothing such as: How often do you drink alcohol? Or do you drink a lot of alcohol? You just do not ask that. You simply do not.” (Care manager, Andersson and Bommelin [32])
	(vi) Care providers reported that due to perceived sensitivity towards the topic of alcohol misuse amongst older people, discussions regarding potential alcohol misuse may only be prompted by concrete evidence - for example, by indicative blood alcohol test results:
	“For me it’s easier if I meet them at hospital, because there there’s nurses and doctors that say when they came to us they had a blood alcohol level of 2.8 and maybe drinking has become a bit too much lately, Then I do not have to bring it up. It makes it much easier. Then you have another opening to discuss the topic.” (Domiciliary care provider, Gunnarsson and Karlsson [36])
	(vii) When discussion of alcohol use was integrated within standard practice, such as within community pharmacy, conveying that this was part of usual care was perceived to minimise negative responses:
	‘Most participants also felt more confident raising the issue of alcohol consumption while undertaking scheduled health checks, when alcohol use could be addressed as simply one risk factor covered in a broader health-related conversation. This minimised client perceptions they were being ‘singled out’.’ (Comment on community pharmacists, Dare et al. [42])
(4) Negotiating responsibility for older adults’ alcohol use	(i) Care providers recognised their older care recipients’ right to self-determination of their own alcohol use. The older individual was therefore seen to be primarily responsible for their own intake:
	“The elderly decide everything in their everyday life and should live a good life, according to the basic values. Within elderly care the right to self-determination principle is applied, regardless of whether or not they have alcohol problems.” (Unit manager, Darwish and Fyrpihl [34])
	(ii) Care providers reported challenges in practice where the older individual lacked insight into their own misuse:
	‘[Domiciliary care managers] describe that these older individuals often lack the insight about their problematic alcohol consumption and few of them ever admit to having any problems. The interviewees find that when the elderly are in denial about their problems it becomes more difficult to help them in the way they need.’ (Comment on domiciliary care managers, Andersson and Bommelin [32])
	(iii) Domiciliary care providers reported particular dilemmas stemming from the older person’s right to self-determination, as they may be expected to play a role in the older person’s access to alcohol:
	“Buying alcohol. Do we? Do not we? Do we have a quantity that you buy? Do you buy for some people and not for others? All those issues. Our view is that the service user has a right and we are not in there to make value judgements about service users and in certain ways we are there as guests of the service user in their home, and there to do obviously what they are not able to. So we would go out and buy. In fact, I’ve just recently had to dismiss a member of staff who had just started—only lasted a week. This member of staff actually refused to buy alcohol for people and to me that is an infringement of the service users’ rights as an individual. So my view was that worker wasn’t able to carry out their full duties.” (Domiciliary care manager, Herring and Thom [30])
	(iv) Care providers reported feeling hindered by inadequate training and support for their roles in addressing older people’s alcohol misuse:
	“There’s not much I can say, [our knowledge and training is] what we brought with us from school, but any other education or information, we just do not have that. But at the same time we need to be prepared for this [addressing alcohol misuse amongst older clients]. When we come across these situations, worries can develop in regards to how we should deal with them.” (Domiciliary care manager, Darwish and Fyrpihl [34])

Uncertainty about legitimacy of drinking as a concern in care provision for older people

Care providers were uncertain about whether older people’s drinking represented a valid issue regarding their care provision. Providers working to support older people with their daily lives, in domiciliary, residential and primary care, recognised that moderate drinking was viewed to be normal and culturally acceptable. Their care recipients’ drinking was usually perceived to warrant little concern in their practice [30–36, 38, 40, 42, 47]. It was evident across providers’ narratives that they held some concern for hazardous drinking amongst older care recipients [32–37, 40, 42, 44]. However, excess drinking was often seen to be less prevalent amongst older people compared with the rest of the population; particularly amongst those working in Sweden and Norway, where a history of temperance meant low-level drinking and abstinence were idealised amongst older people [33, 36, 37, 47]. Those with some focus on addressing drinking in their work recognised how the prevalence of hazardous use was rising within the older population [41, 42] (quote 1i).

The potential for positive as well as negative effects of drinking in older people complicated judgements as to whether alcohol represented a health risk. For providers working to support older people in their daily lives, roles of moderate drinking in care recipients’ lives were seen to contribute positively to quality of life [35, 37, 47]. Domiciliary care providers recognised the social opportunities associated with drinking amongst older clients [33, 36, 37, 46]. Their drinking was accepted and sometimes facilitated within social care organisations, where providers may be involved in supporting clients in purchasing alcohol (quote 1ii). Alcohol was also perceived to play roles in coping with loss of purpose associated with retirement, bereavement and loneliness in later life [31, 33, 40, 42, 44] (quote 1iii). Providers who had cared for individuals for lengthy periods in domiciliary, residential and primary care recognised how this coping role could become excessive and problematic, creating a pathway to alcohol dependence [31, 33, 42, 44] (quote 1iv). The perceived roles of alcohol use in older people’s lives could make it difficult for care providers to intervene when their level of intake was perceived to be hazardous, concerned that alcohol may be “all that is left” in their lives [37, 44].

Rather than being viewed as a legitimate issue for care in its own right, hazardous drinking was framed across care settings as an obstacle to providing care. When providers’ perspectives indicated that care recipients’ drinking was deemed to warrant intervention, it was not in view of preventing harm. Rather, this was usually a result of manifest health consequences or indications of alcohol dependence [30–32, 34–40, 43, 44] (quote 1v).

Providers broadly associated long-term excesses with negative consequences for the older person’s mental and physical well-being [32, 33, 35, 36, 39, 43–45], as well as their self-care [31–39, 44, 45] and social relationships [33, 35, 37, 45]. Amongst domiciliary and residential care providers, shorter term consequences associated with intoxication were emphasised, where older clients could be drinking in the provider’s work environment (quotes 1vi and 1vii). Drunkenness was perceived as having negative effects on behaviour, leading to the care recipient becoming disruptive, aggressive or accident-prone [31, 33, 34, 36–38, 45, 46]. This could threaten the safety of the older person [30–36, 38, 46, 47] and their care providers [34, 38, 46].

Providers working across care settings expressed concerns regarding age-related risks associated with drinking, developed through experiences of supporting older people in particular care contexts [33, 36–40, 42, 44, 45, 47]. Pharmacists and domiciliary and primary care providers involved in medicine provision held concerns about potential interactions of alcohol with medicines [31, 33, 34, 37–39, 42]. Providers working to support older people living in the community recognised the heightened risk of falls amongst older adults consuming alcohol [30, 47]. Providers working in psychiatry and pharmacy with specialist training relating to alcohol were wary of reduced physiological tolerance with age, and alcohol’s effects upon chronic conditions associated with getting older [41, 42]. However, the only examples given of when alcohol was interpreted as a threat was when practitioners were involved in administering medication [34, 42, 45]. In this context, their actions contribute toward interactions between the medication and alcohol which may adversely affect clients’ health. Interventions involved reconsidering medicine provision, rather than addressing the individual’s drinking.

The impact of preconceptions on work with older drinkers

Pre-existing stereotypes of older drinkers shaped care providers’ practice surrounding alcohol use. Across cultures, providers exposed to older people’s drinking through supporting their lives in the community perceived men to be most likely to drink problematically. Consequences of their excessive drinking were most visible, as they often lived alone and tended not to look after themselves [33, 35, 36, 40, 44]. In Sweden and Norway, where historic temperance movements have influenced attitudes towards drinking, alcohol use was perceived to be less common amongst certain groups of older people. These groups included women [33, 35, 36, 40], the oldest old [34, 37, 47] and people in residential care [37, 38].

Preconceptions focussed on characteristics associated with heavier drinking, which proved most problematic and memorable [31–34, 36, 37, 43–45]. Care providers expected that harmful drinking would have visible signs, and looked for these in their practice. Domiciliary and residential carers expected ill health amongst older drinkers from accrued effects of excessive consumption, which reduced independence and created a need for care [31, 33–36, 40, 44, 45]. These expectations guided providers’ exploration of possible excessive drinking [31–34, 36, 39]. Where manifestations of older care recipients’ problematic drinking did not fit their expectations, prejudices could represent a barrier to detection of hazardous alcohol use (quote 2i).

Across professions and cultures, it was a common perception that by later life drinking practices were ingrained [32, 34–36, 39, 42–44] (quote 2ii). For domiciliary carers whose role in intervention was limited (see final theme), when damage was perceived to already be done, this could affect whether they felt seeking support for the individual would be worthwhile [32, 35].

Sensitivity surrounding alcohol use in later life

Excessive drinking was widely perceived to be a morally loaded issue amongst the older age group, because of the common societal expectation not to overconsume [32–34, 36–38, 40, 41, 43–45, 47]. Alcohol was consequently perceived to be a sensitive topic for discussion [30, 32–36, 38, 40–44, 46, 47]. Successful discussion was acknowledged as a factor that could promote an older person’s engagement with appropriate support where drinking had become problematic [32, 34, 43, 44]. Rapport was perceived widely to be an essential prerequisite for successful discussion [30, 32–36, 41, 42, 44–46] (quote 3i). However, rapport was also essential for other aspects of care, and could be threatened by discussion that offends the older person [31, 32, 36, 43]. There were examples where providers recalled this having led to the older person becoming resistant and refusing to accept care (quote 3ii).

Providers working in domiciliary and residential care, who had access to indicators of excessive drinking such as empty bottles, reported that older adults may hide their drinking due to associated stigma [32–34, 37, 40, 41, 43, 44]. This was particularly the case amongst groups where drinking is perceived to be less socially acceptable; for example, home care recipients [33, 34, 37], excessive drinking amongst more socially advantaged individuals [35, 36] and older Scandinavian women [33, 35, 36] (quote 3iii).

Across countries and professions, sensitivity was perceived to be a major barrier to discussing alcohol use, or even recording concerns regarding care recipients’ drinking in notes (quotes 3iv and 3v). Providers were keen to avoid upset [32, 34–36, 38, 40, 43, 46]. Discussions about possible excessive drinking were triggered by concrete evidence [32, 34, 36, 38, 40], such as visible indications of intoxication and overuse, or blood or alcohol use screening results, where misuse cannot be disputed (quote 3vi). In pharmacy settings, where alcohol-related discussion was integrated in consultations, conveying the topic as part of standard practice minimised negative responses [42] (quote 3vii).

Negotiating responsibility for older adults’ alcohol use

Care providers’ emphasis on their older care recipients’ right to self-determine their own drinking was clear [30–39, 45, 47]. Responsibility for decisions surrounding drinking was perceived to lie primarily with the older person (quote 4i). Where the individual lacked insight into risks attached to their intake, this presented challenges in practice across settings (quote 4ii). A lack of insight could stem from poor understanding of the risks of alcohol intake, or inadequate capacity to make decisions, which may result from age-related cognitive decline [32, 33, 35, 36, 40, 41, 43–45]. The older person’s right to self-determination was an underpinning principle guiding domiciliary care, and represented a particular dilemma in providers’ work. There was little scope to act on any concerns about care recipients’ drinking due to limited training and boundaries of their role [30, 32–36, 45]. The provider could often be expected to play a role in the person’s access to alcohol [30, 31, 33, 35, 37, 38, 46] (quote 4iii).

Providers broadly recognised their responsibility in supporting older people with their drinking when it had become cause for concern [32, 34, 35, 38, 39, 41–43, 45–47]. However, their perceived remit in intervening with older people’s drinking related to specific tasks [30–32, 34–37, 39, 40, 42, 43, 45, 46]. Providers universally described how overstretched they were in their work. Discussing drinking could be left aside due to other priorities [30–32, 34, 36, 38, 40–43, 47]. In post-surgical care, stabilising the patient was the goal. Dealing with “chronic” issues like problematic drinking was seen to be the role of primary care [43]. Social care providers discussed how alcohol services focussed on younger people, leaving the older person’s unit to provide for all needs of older clients [44]. Domiciliary carers focussed on supporting the older person to live the life of their choosing, rather than on identifying or addressing problematic drinking. Due to these specific remits, there were repeated examples across care settings where responsibility for identifying and intervening with older people’s hazardous drinking was passed to other providers [32, 34–36, 39, 43, 44] (see Supplementary Figure S1 for diagram of how responsibility was redirected between different providers identified in reporting of included studies). Coordination between care providers was emphasised as important in ensuring the older person was supported to make healthier decisions regarding their alcohol use [31, 32, 34–39, 43–45, 47]. However, this was reportedly difficult to achieve. Where providers did perceive a role in addressing older people’s drinking, they often felt hindered by inadequate training or support for their work [31–35, 37, 38, 40, 42–45, 47] (quote 4iv). Even those with specialist training only felt skilled in specific tasks to support the individual, for example, discussing interactions with medicines in pharmacy [42], and screening for hazardous use in psychiatry [41]. No providers described feeling equipped to motivate or support reduced drinking where it was perceived to be problematic.

Discussion

Health and social care providers’ approach to discussion of alcohol may mean older adults’ needs for support regarding decisions about their drinking are not met. A range of complex factors affected whether alcohol use triggered concern. Social influences, such as cultural norms for drinking and providers’ preconceptions of groups likely to engage in hazardous drinking, influenced judgements. Reservations about the roles of alcohol in older people’s lives, older people’s right to self-determination and sensitivities surrounding the topic of alcohol, all raised questions about the appropriateness of alcohol-related discussion. Care providers’ perceived remit determined how and when they approached alcohol in their practice.

Drinking was perceived as normative, raising uncertainty as to whether it should be addressed in practice. This is concerning, as older adults face increased risk of harm through drinking at levels normalised in many cultures [2]. Providers’ perceptions of who is likely to drink to excess guided their approach. Preconceptions reflected cross-cultural trends for groups likely to use alcohol at higher levels [9], and may raise awareness of aspects of care recipients’ social identity that make them more likely to drink excessively [21, 48, 49]. However, they may present a barrier to older adults in receiving support when they do not meet providers’ expectations [50]. Late-life specific alcohol use screening tools could aid providers’ identification of people who may benefit from support [51].

Societal stigma attached to the topic of alcohol was viewed as a barrier to discussion. Associated reservations surrounding alcohol-related discussion affect providers’ work with care recipients of all ages [21, 52, 53]. Increasing cultural acceptability of alcohol use may make discussion easier to approach [54]. This may particularly affect practice in some Scandinavian countries, where norms are moving beyond lower level drinking [11]. Providers recognised wider roles of alcohol for older adults’ social and emotional well-being; reflecting older people’s holistic view of effects of alcohol [11]. Older adults’ interests therefore influenced providers’ practice. However, this could raise questions regarding the ethics of working to motivate reduced consumption. Respecting older people’s right to self-determination was a common value underpinning providers’ practice, demotivating work to address care recipients’ drinking where concerns were held. This issue was pertinent in domiciliary care, due to their role in supporting care recipients to live the life of their choosing. Reservations for raising alcohol in practice are at odds with the interests of older adults, who expect their care providers to discuss their drinking [55-57]. Standard approaches to support healthier alcohol use incorporate motivational techniques that work with individuals’ perspectives and priorities [58].

Care providers’ experiences and training accrued through their work determined their recognition of risks prompting concern about care recipients’ drinking, and skills for addressing these concerns. Our findings indicated deficits in providers’ capability and opportunity to identify hazardous drinking, and support the older person in making healthier decisions. Capability deficits related to their intervention skills, and knowledge of risks. No provider group demonstrated a complete understanding of risks of drinking in later life. Alcohol-related discussion was rarely a routine task in providers’ practice, limiting opportunity. These deficits are recognised barriers to fulfilling roles in practice [59]. Provider groups had different experiences of older people’s drinking through their work. Those working to support older people living in the community, and with long-term working relationships with care recipients, had insights into roles of alcohol and pathways to dependence. These advised their approach to care recipients’ drinking.

Providers’ perceived remit determined circumstances under which they worked to address alcohol, and their approach. Providers have varying perceived roles and resources to support discussion. Older adults’ complex care needs mean they come into contact with different combinations of providers, and are likely to receive inconsistent support for their drinking [60]. Providers’ role in preventing alcohol-related harm was secondary to their primary job role. The included studies discussed a range of drinking styles, yet providers’ approaches to alcohol were directed by manifest problems including dependence. This suggests providers were not engaging in preventive care; instead looking to provide treatment for those experiencing disease associated with their use. Competing and unclear roles resulted in redirection to other providers, seen to be better equipped to address alcohol. With workloads becoming increasingly onerous in some care systems, this represents a growing concern [61-63].

This novel study applied rigorous, systematic methods to review available literature. Dual-screening and translation of foreign language articles ensure findings represent currently available, relevant material. Through drawing on qualitative literature reporting experiences of care providers, it was possible to present a deeper insight into issues affecting their work than could be gleaned from other types of evidence. This review synthesised evidence from multiple qualitative studies, looking across a variety of health and social care settings. This enabled examination of providers’ different roles and contributions, and associated perspectives. Including studies from across a range of different countries implies that some issues may require global attention.

All included studies were conducted in countries with historical temperance movements, where alcohol is addressed in health policy, and idealisation of restricted drinking may be sustained in residents’ attitudes [11, 64]. The majority of included studies were conducted in Scandinavian countries, where views and use of alcohol are affected by a drinking culture that discourages daily drinking [11, 64]. Some findings may not be generalisable to all high-income countries. Further work is required to deepen understanding of issues affecting care providers’ efforts to address at-risk drinking amongst older people in other countries. This synthesis was limited by the quality of included studies. However, there was sufficient material with which to conduct a meaningful synthesis. Quality limitations were taken into account in formulating reported findings. Triangulating this body of literature through synthesis diminished the baring of individual studies’ limitations on reliability of findings. There is a risk that translation altered the original meaning of included foreign language material. However, this was minimised as translations were conducted by an individual who was bilingual and familiar with the topic.

Providers’ perceptions of older adults’ drinking, and perceived roles in addressing alcohol all affected responses to older care recipients’ alcohol use. These must be addressed to ensure older adults receive appropriate support to meet their needs, including proactive preventive care. Many issues raised within this study could be addressed by training for intervention skills and knowledge of specific risks associated with drinking in old age. Developing approaches must consider the feasibility of care providers’ involvement given their workload, in a climate where care systems are becoming increasingly over-burdened [25].

Click here for additional data file.