Background

The demographics of the world’s population are changing, as people are living longer [1, 2]. The impact on healthcare is pervasive. Older people are frequent users of healthcare services with high multi-morbidly [3, 4]. Healthcare professionals are likely to work with older adults regardless of their speciality or healthcare setting. Increasing the capacity of the healthcare workforce to provide competent and effective care for older adults is an international concern [2]. This includes the need for more specialists in the care of the elderly, such as geriatricians and gerontological nurses [5, 6] as well as generalists with adequate skills in older adult care [2, 7]. Despite this need, working with older adults and the associated specialities is consistently documented as unpopular with healthcare students [8,9,10,11,12,13]. Career preferences formed during training can be predictive of future career choices and behaviour in practice [14, 15]. Therefore, it is important to understand the factors that may influence student preferences for consideration in education and workforce planning.

Previous systematic reviews have explored the factors associated with a low preference of working with older adults in nursing students [16,17,18] and geriatrics in medical students [19]. However, no reviews have included papers of both nursing and medical students allowing direct comparisons to be made. Also, previous reviews have only included preferences for either geriatrics or long-term care settings and excluded studies of educational interventions. Finally, none have included preferences for working with people with dementia where education and care practices are internationally recognised as suboptimal [20,21,22]. This review therefore sought to explore comprehensively the potential factors related to preferences of healthcare students in relation to working with older people and people with dementia using broad criteria for these preferences, comprising any specialities, settings or patient populations related to older people.

Method

A protocol was written adhering to PRISMA-P guidelines [23] and registered on The International Prospective Register of Systematic Reviews-CRD42018104647 [24].

Eligibility criteria

A summary of the eligibilty criteria is presented in Table 1.

Table 1 Summary of Eligibility Criteria
Full size table

Population

The population of interest included medical and nursing students and excluded all other healthcare disciplines. Studies that involved an additional student group were excluded unless findings were separately identified for medical and/or nursing students.

Construct of interest

The construct of interest was student preferences for working with older adults or people with dementia. Due to variability in terms, this included measures of ‘intent to work’, ‘career choices’ or in medical students ‘speciality choice or interest’. There was no restriction on the type of measures of career preferences. Qualitative explorations of preferences were included. The types of preference measured required direct relevance to older adults or dementia. This included preferences measured in relation to patient populations, specialities and settings associated with older adults or dementia.

Types of studies

All empirical articles were included, including quantitative, qualitative and mixed-method studies and theses. Conference reports or opinion pieces were excluded. Studies must have been published in English. Only studies published from 1995 onwards were included. Studies explored career preferences with associated factors or educational interventions. If a study only explored the relationship between an intervention and career preferences, a comparison group was required for inclusion.

Information sources

The initial search was conducted on the 20th September 2018 on the following databases: MEDLINE, PsycINFO, CINHAL, BNI, ERIC and google scholar. To identify further possible relevant articles, the references of included articles and relevant systematic reviews were searched.

Search strategy

Initial search terms were formed during sco** exercises. A specialist librarian was consulted to inform the final search strategy. Key terms included: ((preference adj3 work*)“ “career preference” or “career choice” or “intent* to work” or speciali*ation or “career intent*” or “special*ty choice” or “special*ty interest”) AND (“older adult*” or “older people” or elder* or dementia or geriatric* or aged) AND (student* ad j3 nurs* or “medical student*” or “allied health* student*” or “health* student*). Index terms (e.g MeSH) were also used alongside these key terms. An example search is included in Additional file 1.

Study selection

Identified references were added to EndNote (version X7). After duplicates were removed, articles were screened against eligibility criteria independently by two reviewers (MH & JS) by title and abstract, and then by full text. If there was disagreement, a third reviewer was consulted (SD).

Data extraction and analysis

An extraction template was developed and piloted by the reviewing team (MH, JS, GS, SD) and can be seen in Additional file 2. Only relevant data to preferences was extracted and only data related to nursing or medical students. Statistical probability set at < 0.05, or else recorded as non-significant. For qualitative studies: only major themes (with clear description/quotes), clearly related to preferences, were included as factors.

Papers were extracted independently by two reviewers; MH and either GS and JS, with disagreements resolved by SD. Papers that presented results from the same study were consolidated, to avoid conflation of results.

Quality assessment

Risk of bias was explored using the Mixed Methods Appraisal Tool (MMAT) scale [25], as it takes particular consideration of mixed-method studies [26]. Each paper was rated during extraction by two independent reviewers. Papers were not excluded based on MMAT scores due to the exploratory nature of this review. A narrative description of the overall quality is presented with MMAT scores.

Synthesis

A narrative synthesis approach [27] was used as it allows integration of qualitative, quantitative and mixed-method studies and a quantitative meta-synthesis would not be possible due to the variability of definitions and measurements for career preferences. The use of mixed-method studies means that a data-based convergent synthesis was used [26]. Factors were identified by researchers inductively, using the most consistently used terms by the authors of papers where possible. The labelling of factors for both quantitative variables and qualitative themes were considered with the team to assess fit.

Results

Study selection

Figure 1 below outlines the number excluded at each stage.

Fig. 1
figure 1

Flow Diagram of Study Selection

Full size image

Sixty-six papers were included for data extraction. Four studies were excluded during extraction because they did not meet the inclusion criteria. This comprised 56 unique studies (62 papers) after considering the multiple publication of data. An overview of each study is presented in Additional file 3.

Studies characteristics

The majority of studies were cross-sectional in design (n = 30). There were nine quasi-experimental studies, seven qualitative, seven longitudinal studies and three mixed-method designs.

The highest number of studies was from the USA (n = 14) followed by Australia (n = 6), Canada (n = 5), UK (n = 5) and Israel and China (n = 3). There were two papers each from Hong Kong, Turkey, Taiwan, Sweden, Saudi Arabia, and one from Finland, Ireland, Jordon, New Zealand, Norway, Malaysia, Philippines, Singapore and Sri-Lanka. One paper compared results from Australia and China.

Thirty-eight studies investigated the preferences of nursing students, 17 of medical students, with only a single study exploring both.

Career preference definitions

Of the 18 studies that explored preferences of medical students, 16 investigated interest, willingness or likelihood of pursuing geriatrics, one explored preferences for working with older people and a single study looked at preferences towards working with people with dementia.

The type of career preferences explored for nursing students was varied and was often inconsistent within studies. Only one study investigated the intention to work in dementia care [28]. The majority of studies looked at preferences for working with older people.

Measurements of career preferences

The most common type of quantitative measure of preferences was a single item (n = 23). Eleven studies used a variation of a ranking scale, with the most common based on the work of Stevens and Crouch [29]. Unique scales were described in 14 studies, including one measure of dementia preference [28].

In the seven qualitative studies, methods of investigating preferences included focus groups (n = 4), individual interviews (n = 2) and a semi-structured questionnaire (n = 1). Other qualitative methods used included reflective essays and open text questions.

Research quality

Research quality was variable. Out of a possible score of five on the MMAT, 6 studies scored two, 21 scored three, 22 scored four and 7 scored five. Individual scores are presented in Additional file 3. Consistent issues included the use of non-standardised measures and poor construct definition. Additionally, where educational interventions were evaluated, these often lacked control groups or did not allow for confounding variables. Finally, longitudinal studies had constantly low follow up rates (< 60% for follow up).

Narrative synthesis

A summary of synthesised factors associated with preferences for working with older people or people with dementia can be seen in Additional file 4. Factors are represented by either quantitative variables or qualitative themes. These factors were grouped into seven categories, which are discussed as follows:

  1. 1.

    Student characteristics

  2. 2.

    Experiences of students

  3. 3.

    Course characteristics

  4. 4.

    Career characteristics

  5. 5.

    Patient characteristics

  6. 6.

    Work characteristics

  7. 7.

    Theory of planned behaviour

Student characteristics

Demographics

There was support for a positive association of preference with female gender [9, 30,31,32,33,34]. The relationship with age was limited and inconsistent; a greater preference in younger nursing students was found for working with older people [78].

Theory of planned behaviour

The theory of planned behaviour (TPB) was used as a theoretical model in four studies of nursing preferences; TPB is a model which seeks to explain influencing factors on behaviour [79]. It suggests that people’s behaviour is a rational outcome of considering their ability to perform the behaviour (perceived behavioural control), their beliefs in society and significant others opinions on the behaviour (subjective norms) and individual attitudes to the behaviour. In this reviewed literature, the ‘behaviour’ is a career working with older adults and ‘intention’ is the preference for working with older people. The majority of these studies looked at preferences as the primary outcome rather than behaviour. Only a single study looked at actual behaviour, which was associated with preferences [62]. There was support for attitudes (to behaviour), subjective norms and perceived behavioural control as factors associated with preferences [14, 31, 51, 62].

Discussion

This review has outlined seven categories of potential factors contributing to the preferences of working with older people and provided a comprehensive overview of existing literature for medical and nursing students in this area. This model derived from the literature may have value in understanding healthcare students’ career preferences and designing education to promote work with older adults and people with dementia.

Key findings and implications

The role of undergraduate education

Student preferences for working with older people appear to decrease during training. One explanation is that education has a role in sha** perceptions of the field as low status with an emphasis on technical specialities; this socialisation process is seen as a deterrent for aged care [11, 80]. This has been referred to as a ‘hidden curriculum’ [81]. Key impactful areas during education are clinical placements and educational interventions. The literature suggests these experiences can be key to forming preferences. Nevertheless, the quality of placements, not simply exposure, appears important for promoting professions related to older adults. Examples of quality nursing placement characteristics are found in descriptions of ‘enriched environments’ [64]. These are identified by delivering a sense of security and belonging for the students at the start of their placement, creating purpose and achievement through learning, and reinforcement of the value and significance of gerontology as a profession [64]. The contribution of mentors was also highlighted [52, 66, 67]. Clinical placements and educational interventions should be reviewed to assess the impact upon preferences towards older people. Knowledge on the mechanisms by which placements can influence preferences is limited, but they are suggested to influence via the factors outlined: attitudes; perception of the field; and student preparedness, knowledge and confidence [16, 64] However, robust evaluations of educational interventions in terms of preferences are lacking.

Perceived characteristics of work, patients and career

The characteristics of the work, patients and career, identified in the direct context of influencing preferences, provide insight into why students find working with older people unattractive. This includes the perception of the work as ‘boring’, emotionally challenging, the focus on patient quality of life as opposed to cure as a barrier, the nature of patients’ illness, and communication difficulties, as well as perceived negative aspects of older patients’ disposition. Together this indicates a perception that working with older people and dementia is less valued and challenging. The implication of this is that these perceived barriers may be reduced through education by: establishing the value and improving the profile of work with older people; including the importance and role of healthcare professionals in enhancing quality of life in chronic conditions; and by targeted skill development in perceived areas of difficulty, such as communication and emotional situations. However, while some of these perceptions can be challenged, we must acknowledge the reality of the environmental aspects and career limitations that students recognise. For example, inadequate older peoples services [2] and lack of prestige as described by doctors in geriatrics [82, 83]. Previous authors have suggested how these perceptions explain the unpopularity despite relative high attitudes to older people [31, 73]. Therefore, systemic changes are needed in older people’s services; however, this could be facilitated by inspiring newly qualified healthcare professionals who are able to drive these changes.

Preferences for working with people with dementia

There was a paucity of research in relation to dementia; only two studies explored preferences specifically in relation to working with people with dementia. Potential factors included: female gender; older students; characteristics of the work such as communication and emotional challenges; and educational interventions. Of the factors related to older people, those specifically of relevance include the value of work that appears to stem from the negative perception of chronic and progressive illness and the role of healthcare professionals facilitating quality of life rather than cure. This is pertinent to working with people with dementia.

Two studies mentioned dementia in relation to the importance of exposure to healthy adults to reduce stereotypical prejudices and promote working with older people [31, 78]. The question is how to reconcile this with the evident need for dementia education. A number of new educational programmes are being developed to meet this need [84, 85]. Results suggest the importance of positive clinical experiences and potential for educational interventions to influence preferences positively in dementia-related fields [70]. The implication of these results is that these interventions may offer a way to stimulate interest (both generalist and specialist) but robust evaluations are necessary.

Medical and nursing students

Similar factors were evidenced by both nursing and medical students, specifically, the perceptions of patients and characteristics of work as well as attitudes. The main divergence around was aspects of career pathways leading to differences. There was also more literature on nursing students with more diversity in the types of nursing preferences explored, this is likely due to the nursing career paths being relatively unstructured comparatively to medicine. The inclusion of both medical and nursing students is a strength of this review as this is a multi-professional issue and allows preferences to be view in this context of both wider policy and education.

Future work

Significant gaps in research include an exploration of positive factors, longitudinal data, validated preferences measures and clear definitions of preferences. There is also a paucity of robust evaluations of education interventions and understanding of mechanisms of influence. The development of conceptual frameworks would be critical in hel** to conceptualise these factors and the relationships between them. One clear area for future research is preferences related to dementia.

Limitations

Studies were not excluded based on quality, which could have introduced bias into the review [25]. However, this was an exploratory review looking at possible factors with the aim to be comprehensive. Secondly, we have grouped different types of preferences together, although many studies did not define either what they meant by working with older adults or equate particular settings with working with older adults. Future work should provide definitions, including considering interpretations of student responders. Finally, the analysis was restricted to a narrative synthesis and therefore the magnitude of associations was not examined, giving no indication on the relative weighting of factors. Furthermore, we did not make a distinction between univariate links and multivariate analysis; given that many of the studies explored preferences not as the primary outcome, with multiple correlations being presented, there is the risk of type-1 error. However, this review has three main strengths: a rigorous systematic review methodology; it is novel in its inclusion of medical and nursing students in older adult’s preferences; and it is the first to explore preferences for working with people with dementia.

Conclusion

Seven overall categories of factors were found and provide implications for education to promote working with older people. It was found that while there is a wide and varied literature relating to older adults, understanding of factors associated with working with dementia specifically is limited and is a key area for future research.