METHOD
The study team comprised a health services researcher, two academic GPs, a clinical epidemiologist/public health specialist, and an academic nurse. A thematic synthesis approach was used from a constructivist perspective. This aimed to produce outputs directly relevant to policymakers and practitioners, and allows reviewers to look for differences in perspectives according to study characteristics, such as type of HCP.22,23 The protocol was registered on PROSPERO (reference: 42017055207).
How this fits in
Older people are often prescribed antidepressants and are less likely than younger adults to be referred to psychological therapies, particularly when they are aged ≥80 years. Qualitative research was synthesised to understand how healthcare professionals (HCPs) manage late-life depression, particularly regarding psychological therapy referrals. It was found that clinicians had little time to negotiate the complex issue of depression and so prioritised physical over mental health needs in older people. HCPs reported a lack of treatments that were both available and appropriate, and so management depended on the individual practitioners’ skills rather than a coherent structure.
MEDLINE (1946–March 2018), Embase (1974–March 2018), PsycINFO (1806–March 2018), CINAHL (1937–March 2018), and the Science Social Sciences Citation Index (SSCI) (1900–March 2018) databases were searched (database search terms are available from the authors on request). Grey literature was identified through searches of the e-theses online service (EThOS) (inception–April 2018). Studies were eligible for inclusion if:
most participants were primary or secondary HCPs;
qualitative methods were used to collect and analyse data in a substantial part of the study; and
they featured views and experiences of the treatment and management of older people with depression.
Studies were excluded if they:
focused on depression management in younger/all age groups or in people with a specific medical condition, for example, post stroke;
explored the effects or implementation of new interventions;
studied later-life mental health without specific depression data;
included views of social services, third-sector, or trainee staff only;
focused on pharmacotherapy or suicide only;
were quantitative or non-empirical; or
were published in a language other than English (these were excluded due to a lack of translation facilities).
Two reviewers independently assessed 10% of the titles and abstracts (88% agreement, with disagreements resolved through discussion), and a further 10% (91% agreement). Each reviewer then screened half of the remaining records. Full texts were appraised independently by the same two reviewers, with disagreements resolved through discussion or consultation with the whole team.
Data extraction and synthesis
Data relating to study aims, location, participants, data collection, analysis, themes, and authors’ main implications were extracted by one reviewer and papers were imported into NVivo (version 12) for synthesis. Study quality was appraised using seven questions derived from the Critical Appraisal Skills Programme checklist and other checklists.24–26 Two reviewers independently assessed study quality according to individual items and gave an overall subjective judgement of:
quality (very poor, poor, not very good, good, very good, excellent); and
reporting quality (poor, acceptable, good).
Quality was assessed to provide an overall summary of the evidence base, but studies were not excluded and findings were not weighted within the synthesis on the basis of quality; this was because the role of quality assessment within qualitative systematic reviews has a number of associated debates, as highlighted by Dixon-Woods et al.26
A thematic synthesis approach was followed, which comprised coding text, developing descriptive themes, and ‘going beyond’ the primary studies to develop analytical themes and answer the questions posed by the review, in line with recommendations by Thomas et al.22 Results sections of included papers were coded line by line by one reviewer, with independent analysis of one-third of the papers undertaken by another. These codes were aggregated to create descriptive themes that were summarised and discussed by all authors (the framework is available from the authors on request).
In order to go ‘beyond’ the data, analytical themes were developed by one reviewer. Potential connections and groupings were modelled, with statements written out hypothesising links, connections, and themes. These were compared with coded data within and across studies and HCPs, and refined until subthemes and themes were constructed. The analytical themes were reviewed by all authors, refined, and then agreed on.
RESULTS
Out of 1471 unique records, 161 full texts were screened and 27 studies were included in the qualitative synthesis (n = 2 arising from the same dataset) (Figure 1). The majority of studies were carried out in Western countries (eight in the UK, eight in the US, five in Australasia, three in Scandinavia, and one in Canada), plus one each in Taiwan and India (Table 1); detailed study summaries are available from the authors on request. Studies reflected both publicly funded and insurance-based systems. Qualitative data were mostly collected via interviews18,19,27–42 and/or focus groups34,39,43–46 with two ethnographic studies,47,48 one conference and nominal group technique,49 one mixed-methods survey,50 and one multiple case study.51 Two-thirds of studies were of good/very good overall quality (Table 1); most met the checklist criteria and the vast majority were well reported (Table 2).
Figure 1. Flow chart of search and appraisal process. HCP = healthcare professional.
Table 1. Study characteristics.
Table 2. Quality appraisal for included research papers, n = 27
Most HCPs were sampled from primary and community healthcare (for example GPs, practice nurses, or home health nurses),18,19,28,29,31,32,34,39–42,44,47,48,50 with six studies sampling both primary and secondary care professionals,27,35,45,46,49,51 and a small number studying HCPs in care settings.33,38,43 One study sampled community psychiatric nurses (CPNs)36 and one included practice counsellors.18 As similar groups of professionals were often referred to with different names in different countries, each professional was grouped under a UK heading (for example, GPs for primary care physicians) in the thematic synthesis.
Five themes relating to management were identified:
avoidance of medicalisation of social circumstances;
assumptions regarding older people and mental health;
prioritisation of physical health across healthcare settings;
the ‘postcode lottery’ of therapeutic options; and
variation in skills, training, and approaches across all settings.
Avoidance of medicalisation of social circumstances
Late-life depression was felt to lack suitable therapeutic solutions as it was considered to mainly arise from ‘justifiable’ causes, many of which related to ageing. The majority of HCPs across all countries primarily attributed late-life depression to difficult social circumstances and, in particular, age-related social issues (for example, loneliness or bereavements) and/or physical health issues, frailty, and functional decline:18,19,32–34,37,39–44,51
‘GPs described depression as part of a spectrum including loneliness, lack of social network, reduction in function, and very much saw depression as “understandable” and “justifiable”.’19
Many GPs and nurses, therefore, felt there was a definite difference between sadness or distress that ‘understandably’ related to these issues, and clinical depression, but rarely defined where this border lay.18,19,31,32,39,43 Consequently, across all studies discussing this issue, there was a clear tension as to whether medical treatment (particularly antidepressants) could be beneficial or represented a medicalisation of social issues. This was further complicated by widespread views from GPs and nurses that addressing depression in some way was essential:18,19,27,31,32,39–41,43
‘GPs tended to acknowledge social and emotional causes that required non-drug interventions they could not always provide and, although antidepressants offered a solution to some patient’s problems, there appeared to be sense of unease about prescribing a medical intervention for a social cause.’31
‘Social solutions’41 (for example, day centres) were considered the most-appropriate approach;19,34,35,39,41,44,45 however, although these addressed the perceived cause, they were not always regarded as effective, leading to therapeutic pessimism:
‘ ...“a man who is clearly isolated ... taking him out to the day centre, that’lI be good for him won’t it, sit in a room with other demented, depressed people, and make your mood lift. It’s not gonna happen.”
(G3, psychologist)’51
When depression was conceptualised as a response to physical illness and/or disability,18,19,32–34,37,40–42,51 even fewer solutions were identified, with the person’s future regarded as negative.51 Physical health problems were deemed to be a barrier to psychological treatments by psychologists and CPNs,36,51 and psychological approaches were seen as inappropriate by some GPs.41
‘ ...“usually because they’re bloody sick, and is a psychologist going to help that? I don’t think so.”
(D6)’40
A minority of home care nurses and GPs reported using clinician support to encourage adaptation to disability;32,34 otherwise, in response to disability, medicalisation (and subsequent antidepressant prescribing) was seen as a better alternative to doing nothing.31,40
Assumptions regarding older people and mental health
HCPs held a number of assumptions regarding older people’s attitudes to depression; the most pervasive was that older people normalised depression as part of ageing, isolation, and decline, and felt it to be stigmatising.18,19,31,34,37,38,40–42,46,51 Few discussed the idea that older people may have had early or mid-life experiences of depression. HCPs, consequently, assumed that late-life depression was likely to be hidden; that older people were resistant to articulating depression or distress, instead ‘sprucing up’ for the GP or presenting somatically:27,28,31–34,37,39–42,49,51
‘Older people were reported to attribute symptoms differently, and to have more rigid and strongly held beliefs about stigma, the desirability of coping unsupported, and the implications of failure to do so.’ 41
GPs and district nurses, therefore, felt that depression took time, effort, and skill to actively look for by indirectly focusing on symptoms and related concepts such as loneliness or homesickness, using screening tools, or medicalising language (for example, ‘clinical depression’ or ‘neurobiology’) to reduce stigma.19,27,30,32,37
This assumption of stigma and ‘hidden’ depression did not translate to home and residential care settings. Observational and HCP-reported data suggested overt symptoms, such as crying, poor self-care, and reporting feeling sad,33,39,47 were displayed, although this did not increase the likelihood of treatment:
‘ [observation] A patient says “I’m feeling sad today,” while the nurse comments on a blood pressure reading.’47
Although individual treatment preference was considered more important than age in some studies,32,34,39 there were widespread assumptions that older adults disliked, and were reluctant to engage with, any mental health-related treatment. Psychiatry was considered particularly stigmatised, and so psychiatry referrals were a last resort:41
‘Attaching depression to mental illness was also reported as a barrier to older adults’ seeking mental healthcare services […] “They will not accept seeing a psychiatrist because, in general, people believe that psychiatry is for treating crazy people”.’37
Decision-making power in late-life depression rested chiefly with professionals. Some HCPs (mainly GPs) assumed that older adults were uninterested in talking therapies or that they would be ineffective,19,31,32,35,37,46 particularly if they were computer based (such as online cognitive behavioural therapy),42 and so were likely to dismiss these as an option.35 This did not always preclude referrals if other treatments were ineffective, but ageist stereotypes were also evident in psychological therapists and CPNs, who felt that older people were unwilling to change — ‘“… it would not be obvious to me at 76 what would improve.” (G3, psychologist)’ 51 — and discharged themselves more quickly.36,51
Some HCPs felt that older adults conceptualised antidepressants as having a stigma or being addictive, and so required persuasion to take them.28,31,37,40 Despite GPs feeling they had greater influence on older, as opposed to younger, people,34,41 their assumptions that older people were resistant meant they felt that it was easier to circumvent depression during treatment by prescribing antidepressants for ‘insomnia’ or ‘pain’, or using GP support:32,37
‘I would say that even seeing someone and talking a bit in the GP surgery is a treatment in a sense although they might not think of it like that, they might just think it’s a chat.’
(GP9, p3)’ 32
Prioritising physical health across healthcare settings
Implicitly and explicitly, physical health issues were prioritised over mental health. 19,27,28,32–37,39,41,42,46,47,51 Severe depressive symptoms could prompt action, but severity was usually defined in terms of physical impact (for example, suicidal ideation, impact on discharge planning).34,41,46 The subject of depression was, therefore, sometimes avoided completely, despite some recognition that physical and mental health interacted:32,34,45
‘[observation] A patient says “I just want to die,” and the nurse nods head without verbal response and asks if the patient has had a recurrence of a bothersome physical symptom”.’ 47
This was also affected by organisational time pressures: non-psychiatric secondary care was considered a poor place for depression management, due to multiple assessments from varied professionals, the acute focus, and the lack of an identified responsible person and follow-up.46 The widespread view that late-life depression was best managed in primary care in many UK, Australian, US, and Taiwanese studies,19,37,39–41,46 with mental health services as consultative support,41 was, however, at odds with the time available to GPs:
‘ [They are] commonly complaining of not having enough time to address the many complex issues surrounding depression in later life. “In 10 minutes there is a lack of time as to what you can do with somebody … sometimes you don’t get to the nitty gritty.”
(GP5, p4)’ 32
In the US, home nursing visits were only eligible for insurance reimbursement if the older person had a documented need for physical health care,34,36 while UK district nurses’ time was limited more implicitly.40 Mental health was viewed by many HCPs as outside of their role, which was reflected in their lack of mental health training.19,29,32–34,39,40 In this case, mental health was addressed only if they had time or could do so outside of appointments; for example, GPs with extended consultation times32 or home care nurses scheduling evening appointments:34,37,39
‘The district nurses described strategies to provide the time they felt people needed but they were unable to provide as an accepted part of their role. This involved logistical approaches such as leaving certain visits until the end of the day or scheduling a visit at weekends.’ 39
Addressing psychological issues as a key part of physical health care for a condition (for example, cancer) was viewed as more accessible for older people.42 Some GPs justified a physical focus as older people were at higher risk of illnesses such as dementia or cancer, which could have a similar presentation.32,37 However, many felt that, despite the commonality of depression in people who are frail and housebound across primary, community, and acute settings, it was much more likely to be overlooked:31,32,41,46
‘Paradoxically, as old age and ill health became more integrally associated with depression and its treatment, the latter was less often mentioned in consultations.’ 31
Conversely to this prioritisation of physical health, across the studies, few concerns were expressed regarding how antidepressants might impact on physical health issues (for example, increased falls risk).19,40 A sense of therapeutic inertia seemed to occur once patients who were frail were taking antidepressants as clinicians feared upsetting a delicate equilibrium or giving rise to care problems.31,33,43
The ‘postcode lottery’ of therapeutic options
Although psychological therapies, psychiatric services, social workers, or social activities were considered appropriate to the perceived causes of late-life depression, they were constrained by wide differences in provision across localities.19,28,31–35,37,39–42,44,46,48,50 Long waiting times, narrow eligibility criteria, poor integration with other care, being inappropriate to needs or preferences, financial constraints, and limited duration of support were key issues.31,34,39–41,43,46,50 This led to an automatic discounting of psychological therapies or social approaches as an option:
‘“You’ve got to be pretty sick or mad to get any extra help ...’”
(PN1)’ 19
As GPs and nurses felt that depression did need to be addressed when raised, having solutions that were both appropriate and available led to a greater inclination to identify late-life depression, regardless of other factors; particularly if HCPs were less confident about managing it themselves. However, the reverse was true when services were not available:19,32,42
‘The majority of healthcare professionals described a reluctance to make the diagnosis of depression in an elderly person because of a feeling that they had nothing to offer the patient.’ 19
When nothing else could be offered, GPs and community nurses tended to provide support themselves to the older person in various ways or prescribe antidepressants.31,43 Limited referral options were not always related solely to provision, however, some clinicians reported needing further local service knowledge.33,46,49
Variation in skills, training, and approaches across all settings
Differences in HCPs’ skills, interest, and perceived role in depression management were reported across all settings. Those with greater confidence in managing depression (usually GPs) were more likely to raise the topic.19,30,32,34,40–42,45 Confidence related to training and experience, something for which non-psychiatric nurses and acute care professionals expressed a need.19,29,34,35,39,42,46
Personal interest also influenced individual GP approaches, for which a number of studies developed typologies.28,32,48 For many GPs, active listening and using the therapeutic relationship to change views about depression was seen as an effective and sufficient treatment strategy, especially for mild symptoms,19,32,37,39–41,44,45,48 and so they were reluctant to refer on. The minority who felt insufficiently trained to offer this,19,41 or felt that their remit was only to refer on or prescribe medication,28,32,51 were more likely to refer when these services were available.
Different approaches and attitudes had clear effects on patients’ and teams’ experiences:
‘Residential aged care services that had positive experiences with GPs found the referral process to other external services for depression and subsequent outcomes for residents far more positive and beneficial. On the other hand, an equal number of participants expressed disappointment at the services provided by GPs.’ 33
Some nurses felt confident they had a role in depression identification and management by reporting concerns to the GP and coaching patients on broaching depression,27,30 emotionally connecting with patients,47 referring to local services,33 or counselling regarding physical loss and disability.28 However, some nurses and many GPs felt it was outside of the nurse’s role and that these HCPs lacked skills and/or training to manage it:19,29,33,39,42,43
‘Nurse: “We are not skilled in differentiating between these conditions. If they cry, we call it depression and give them antidepressants. And that’s it”.’ 43
Interprofessional communication played a key role in home nursing and residential care settings.28,37,39,42,43,47 The greater number of communication channels required to refer to primary care and the associated hierarchical issues meant that concerns could be lost or dismissed, even if depression was adequately identified.33,38,39,43
Other relevant professionals were thought to include social workers (US and Australia),27,46 psychiatrists,45 and multipurpose community health workers (India).44 Psychologists and psychiatrists were considered skilled in late-life depression management, but little further information regarding psychiatrists’ views was found. Nevertheless, strong mental health service collaborations were considered important and increased other HCPs’ confidence in managing depression; these collaborations appeared to be incidental, however, arising from interested individuals rather than a clear structure:32,41,42,49
‘… most participants emphasised that their best collaborations evolved on a case by case basis as they found health professionals on mental health teams with whom they could readily consult, solve problem, and share information.’ 49
DISCUSSION
Summary
A total of 27 qualitative studies of HCPs’ management of older people with depression were systematically reviewed. Decisions regarding the identification and management of depression in later life were underpinned by strong assumptions across all settings that older people were resistant to discussing depression and its treatment; these views were compounded by physical health being prioritised over mental health and a high variation in skills and training among HCPs, particularly for nurses. Beliefs about the causes of depression and its social origins underpinned decisions regarding when and what treatment was appropriate, particularly for GPs, which further depended upon a ‘postcode lottery’ regarding the treatments that were available for consideration.
Strengths and limitations
Studies were systematically identified and drew on views from a range of professionals, countries, settings, and ethnicities. This provided a more complete picture of how depression is managed in older adults, including in those with frailty or multimorbidity.
A constructivist approach was taken to identify and contrast multiple conceptualisations of late-life depression. Other reviews have focused mainly on GPs,21,52 but the one presented here also included the views of nurses and other HCPs. However, it is important to note that most studies were carried out in high-income Western countries and, due to a lack of resources for translation, only English-language studies could be included. One excluded study, reported in German, found themes almost identical to the review presented here.53 Although family played a larger role in the studies from Taiwan and India, similar themes were also identified, so the findings presented here may have some transferability to these settings.
Meta-ethnography, which may have offered greater conceptual integration, was not used but the ability to compare across HCPs — as outlined by Barnett-Page and Thomas23 — was considered a key advantage for this review. A further strength is that the majority of included studies were of good quality. Theses also provided a richer source of data than published papers but, as it was only possible to access UK theses, it could be that further qualitative evidence exists from other countries.
Comparison with existing literature
Similar themes of depression being considered a normal response to challenging social circumstances, caution being exercised regarding medicalising social issues, and a lack of psychological therapies have been found in HCPs’ views of depression in adults.21,52 However, ideas such as ‘secondary gains’ to a depression diagnosis (for example, avoiding social problems, feelings of powerlessness, or work) or overtly negative attitudes were not discussed regarding late-life depression,21,52 possibly as late-life depression was felt to be more justifiable.
Other elements have been discussed elsewhere that were not discussed in this review, including taking a short-term view for older adults (for example, GPs’ reduced concern around addiction to benzodiapezines as people were unlikely to live for much longer54) and involving family and/or supporting carers in depression management.55,56
Social workers and family carers expressed similar views around late-life depression, a lack of priority for mental health, and low availability of resources.57,58 Poor access to psychological therapy for older adults has also been documented elsewhere;7 this may be compounded by ageist views and a lack of trainees in psychological fields who are motivated to work in the sector.59
It has also been identified elsewhere that older adults normalise depression, although some studies suggested that older people were more likely to raise depression in emotional rather than somatic terms and lacked awareness of psychological treatments,60 despite reporting preferences for them over antidepressants.11,12,14
Implications for research and practice
This review suggests that primary care services for older people do not currently prioritise older adults’ mental health to the same extent as their physical health, which is compounded by a lack of referral options suitable to older people’s needs. Further investment in psychological and social resources is needed to enable mental health in later life to achieve equitable priority with physical health, particularly if older adults are to be encouraged to use psychological services.
UK guidelines recommend planning partnerships between local authorities, the NHS, community organisations, and voluntary-sector providers to improve mental wellbeing and promote independence in older adults.61 Examples of successful management identified in this review suggested that staff training and improved links to other services could result in older people being better supported. Within all services, the roles and responsibilities of HCPs — and nurses in particular — need to be outlined more clearly.
Most GPs whose views were reflected in this review felt that late-life depression was within their remit. Patient views have suggested that some feel that GPs would not be receptive to discussing mood, and that having a person outside of the GP consultation was beneficial.19,62
Research into GPs’ views of late-life depression has received substantial coverage and replicating this further in high-income countries is likely to be unnecessary. However, as there was a notable paucity of views of psychological therapists or psychiatrists in this review, despite their role in treating late-life depression, this remains an area for further qualitative research.
Internet or bibliotherapy psychological approaches were also rarely discussed, and, although depression in populations with frailty was considered common, fewer solutions were identified for this subpopulation. Further research into effective and equitable treatments for late-life depression is needed.
In conclusion, this systematic review of qualitative studies suggests that depression in later life can be managed in primary care, but needs to be given greater priority to address the complex needs of older adults. Sufficient staff training and clarity of staff roles is also required. Investment in psychological therapies that are suitable for older adults, along with other social referral options, are needed to facilitate the identification and treatment of late-life depression.