What works, why and how? A scoping review and logic model of rural clinical placements for allied health students

Background Allied health services are core to the improvement in health outcomes for remote and rural residents. Substantial infrastructure has been put into place to facilitate rural work-ready allied health practitioners, yet it is difficult to understand or measure how successful this is and how it is facilitated. Methods A scoping review and thematic synthesis of the literature using program logic was undertaken to identify and describe the contexts, mechanisms and outcomes of successful models of rural clinical placements for allied health students. This involved all empirical literature examining models of regional, rural and remote clinical placements for allied health students between 1995 and 2019. Results A total of 292 articles were identified; however, after removal of duplicates and article screening, 18 were included in the final synthesis. Australian papers dominated the evidence base (n = 11). Drivers for rural allied health clinical placements include: attracting allied health students to the rural workforce; increasing the number of allied health clinical placements available; exposing students to and providing skills in rural and interprofessional practice; and improving access to allied health services in rural areas. Depending on the placement model, a number of key mechanisms were identified that facilitated realisation of these drivers and therefore the success of the model. These included: support for students; engagement, consultation and partnership with key stakeholders and organisations; and regional coordination, infrastructure and support. Placement success was measured in terms of student, rural, community and/or program outcomes. Although the strength and quality of the evidence was found to be low, there is a trend for placements to be more successful when the driver for the placement is specifically reflected in the structure of the placement model and outcomes measured. This was seen most effectively in placement models that were driven by the need to meet rural community needs and upskill students in interprofessional rural practice. Conclusion This study identifies the factors that can be manipulated to ensure more successful models of allied health rural clinical placements and provides an evidence based framework for improved planning and evaluation.


Background
Most health workers live and work in cities, yet almost half of the world's population currently live in rural and remote areas [1]. Challenges attracting and retaining a full complement of health workers in rural and remote communities is widely recognised as being a significant contributor to rural residents experiencing poorer health outcomes than their metropolitan counterparts [1,2].
The World Health Organisation (WHO) recommends a number of ways to address the issue of under-supply of health professionals in rural areas, including national policy, regulatory interventions, financial incentives, personal and professional support and the education of health students [2]. This research focuses specifically on Australia's rural allied health (AH) workforce pipeline, while acknowledging there are also similar and different issues and incentives for medical and nursing workforces.
While there is no universally accepted understanding of what professions are considered part of the AH workforce, it is generally understood as not including the medical, nursing or dental professions. "Allied health professionals are university qualified practitioners with specialised expertise in preventing, diagnosing and treating a range of conditions and illnesses. Allied health practitioners often work within a multidisciplinary health team to provide specialised support for different patient needs" [3] . In Australia, one state has recently recognised 27 allied health professions and has identified them as either therapy focussed professions (such as physiotherapy, occupational therapy, speech pathology, podiatry, dietetics, social work, psychology, exercise physiology and more) or science focussed professions (medical laboratory science, radiology, nuclear medicine, orthoptics, pharmacy, sonography and more) [4].
In support of WHO recommendations, countries like Australia have invested heavily in closing the ruralmetropolitan health gap. This has been done through increasing the number of health students and, more specifically, investing in educating students closer to rural communities, bringing students to rural communities, and matching curricula with rural health needs [2].
Significant funding has been directed towards University Departments of Rural Health (UDRHs), of which there are currently 16 located across rural Australia [5]. The UDRH program aims to provide education and training facilities in non-metropolitan centres with the aim of helping attract health professionals to practise in rural and remote communities [6]. In 2013, an estimated 18% of annual university enrolments in 10 leading health disciplines accessed UDRH clinical placements [7]. The capacity for the current education system to accommodate the minimum clinical training or work integrated learning hours required for course accreditation and subsequent professional registration for a (now) large supply of health students, however, is severely limited [8]. In particular, the growth in new courses across AH in Australia means that there is now extreme competition for access to clinical placements in all settings [8].
Rural clinical placements (RCPs) or rural work integrated learning opportunities [in this article the term RCPs will be used] for AH students provide a number of benefits, including: a clinical placement opportunity to meet course accreditation and professional registration requirements; developing skills for (rural) practice; influencing consideration of rural employment; and providing unique learning opportunities [9][10][11][12]. Therefore, RCPs offer a potential solution to the placement difficulties for universities whilst simultaneously helping to address broader issues of rural health workforce supply and rural readiness to practise [12]. Often RCPs for health students attempt to achieve all these goals, thus creating a complex set of circumstances where it is difficult to understand or measure how success is facilitated.
There is, however, a paucity of well-synthesised evidence that reflects the complexity of the RCP environment for AH students, in particular why and how particular models of RCP work [2] and how these models impact student, service, patient and key workforce outcomes [13]. It has been argued that traditional systematic reviews, which impose a strict hierarchy of evidence, rarely reflect the complexity of the context in which the interventions are operationalised [14][15][16]. As such, there is a growing argument that a fuller synthesis of 'relationships, mechanisms and meaning' within the evidence base is required by managers and policy makers [17].

Methods
Aim, design and setting of the study Given the complexity of factors that can influence the development, implementation and outcomes of RCPs [12,[18][19][20][21][22][23][24][25][26], this research is underpinned by a logic framework [27] to identify and illustrate how different elements of RCPs and associations between elements may impact on the outcomes and ultimately the 'success' of RCPs.
A preliminary search of PROSPERO, the Cochrane Database of Systematic Reviews and the JBI Database of Systematic Reviews and Implementation Reports was conducted, and no current or in progress systematic reviews on the topic were identified. The objectives of this research are to identify different models of RCP for AH students; to better understand the drivers, contexts, mechanisms and outcomes of these models; and how these elements come together and interact to influence the 'success' of the RCP. The end goal of this research is to provide universities, UDRHs and placement sites with clarity around the elements of RCPs they could strengthen according to the outcome they wish to influence. The findings are synthesised using a logic model to identify a guiding framework that can be applied across a range of contexts for the development of sustainable, quality RCPs.
The scoping review was conducted in accordance with the Joanna Briggs Institute methodology for systematic scoping reviews [28] using the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) Checklist.

Review questions
Using a realist perspective [27], the overarching review question is: what AH RCP models currently exist for AH students and what AH RCP models are more successful?
To answer this question, the following sub-questions were developed: -What are the key drivers of (or needs underpinning) regional, rural and remote clinical training placements for AH students?
-In what types of contexts do AH RCPs take place (e.g. setting, staffing, organisation, structure)? -What mechanisms (barriers and facilitators) are required for successful AH RCPs? -What success measures have been used to capture the impact or effectiveness of AH RCPs? -What is the relationship between drivers, contexts, mechanisms and outcomes?

Participants
The review considered studies that included AH students and those on interprofessional (IP) placements with other non-AH disciplines (such as medicine or nursing). The search did not include clinical placement studies concerning only nursing or only medical students. A broad definition of allied health was used, with professional titles taken from Services for Australian Rural and Remote Allied Health (SARRAH), the AH portfolio of the New South Wales Health Education Training Institute, the Victorian Department of Health and Human Services, AH workforce and the Allied Health Professions Australia (AHPA) websites (see Table 1 for all terms used). Type of study All, quantitative research (randomised controlled trial, controlled clinical trial, controlled before and after study, uncontrolled before and after study), qualitative (grounded theory, ethnography, action research, exploratory approaches, phenomenology, and systematic reviews).

Approaches
Subject searching, citation searching, contact with authors.
Range of years January 1995-May 2019.
Limits English, human.
Inclusion and exclusions b Inclusion: empirical study of an intervention aimed at allied health c student clinical placements undertaken in regional, remote and/or rural areas. Exclusions: developing country health care, non-empirical research (grey literature, commentary, editorial, discussion piece), conference abstracts, not allied health (medicine, nursing, dental), not rural/remote/regional, not clinical placement interventions or models, not theses/dissertations.

Concept
The review considered studies that explored models of clinical placement. It did not consider studies that tracked longitudinal rural practice intentions of AH practitioners (AHPs) as these studies cannot be attributed to a single particular rural placement model.

Context
The review considered studies that were undertaken in regional, rural or remote contexts in Australia and in other developed countries, as defined by the United Nations' World Economic Situation and Prospects country classification [30]. It did not consider studies from metropolitan or urban contexts or from developing countries.

Types of sources
The review considers both experimental and quasiexperimental study designs, observational studies, qualitative studies and systematic reviews but not theses, dissertations or grey literature. Studies published in English since 1995 were included. Table 1 describes the search strategy and inclusion criteria in full.

Search strategy
The search strategy targeted published studies. An initial limited search of MEDLINE and CINAHL was undertaken to identify articles on the topic. The text words contained in the titles and abstracts of relevant articles and the index terms used to describe the articles were used to develop a full search strategy ( Table 1). The search strategy, including all identified keywords and index terms, was then adapted for each included information source. Reference lists were not screened for additional studies. The databases searched included: Academic Search Premier; CINAHL; EMCARE; InfoRMIT:Health Collection; MEDLINE and ProQuest. Other unpublished studies, research reports and grey literature were not used for this review.

Study selection
Following the searches, all identified citations were collated and uploaded into EndNote version 7 and duplicates removed. Titles and abstracts were screened by four independent reviewers for assessment against the inclusion criteria for the review (AM, SN, RM, CC; see Table 2). Potentially relevant studies were then retrieved in full and assessed in detail against the inclusion criteria by four independent reviewers (AM, CC, SN, RM). Reasons for exclusion of full text studies that did not meet the inclusion criteria were recorded on the preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement [31]. The results of the search are presented in a PRISMA flow diagram (see Fig. 1).

Data extraction and synthesis
Data were extracted by four independent reviewers (AM, SN, RM, CC) using a logic framework. The data extracted included specific details about the population, concept, context, study methods and key findings relevant to the review objective using pre-defined program logic headings: 'drivers' (or needs underpinning), 'context' (setting, staffing, organisation, structure of RCPs), 'mechanisms' (barriers and facilitators) and 'outcomes. The data extraction strategy was not modified during the process of extracting data from each included study. Papers were appraised using Joanna Briggs Institute (JBI) critical appraisal tools and the mixed methods assessment tool (MMAT) [32]. Data were synthesised using thematic analysis with the final logic model synthesised using colour coded charting of themes across the logic headings (Additional file 1).

Data presentation
Extracted data is presented in tabular form with narrative used to describe findings using program logic headings (Additional file 1).

Results
In total, 292 articles were identified. After removal of duplicates and article screening, 69 papers were considered for inclusion, with 18 included in the final synthesis ( Fig. 1 PRISMA). Australian papers dominated the evidence base (n = 11) ( Table 3 and Additional file 1). The dataset(s) supporting the conclusions of this article is (are) included within the article (and its additional file(s).
What are the key drivers of rural clinical placements for allied health students?
Macropolicy level drivers The evidence base identified macro (policy) level drivers aimed at increasing the size of Australia's rural allied health workforce to address issues relating to rural health inequality and underservicing of rural areas [33,34,36,37,41,43,45,47,49]. As such, the primary macro level driver for RCP identified in the literature concerned the need for attracting AH students to rural health employment upon graduation. A smaller driver was identified which related to the capping of university places and increasing the number of student placement opportunities or placement capacity [37].

Mesouniversity level drivers
A key driver within the university sector for innovation in RCPs and increasing access to more RCP opportunities is    [50], p52. Equally, the literature demonstrates that within the university sector, the provision of RCPs is driven by a commitment to increasing the supply for the rural AH workforce [33-37, 40, 44] and, related to this, ensuring graduates are work-ready for rural employment [38,43]. For example Wolfgang et al. state 'Creating positive rural experiences for occupational therapy students on placement could potentially improve the recruitment and retention of occupational therapists in rural and remote areas … and influence occupational therapy students' decisions to work rurally.' [37], p204. Universities were also driven by a commitment to improving access to AH services in rural areas through student clinics or student provision of services whilst on placement [38][39][40][41][42][43][44]50]. Allan et al. for example describe how university clinics were proposed as one way to increase the number of clinical placements available for AH students, while simultaneously providing healthcare to rural communities [38]. . Further, the university sector is also driven to supply RCPs as a unique learning opportunity for students where they can learn particular skills that are key to student competency, such as interprofessional practice [35,[45][46][47][48][49] or cultural competence in working with minority and/or vulnerable population groups. Gum et al. assert that 'Rural communities provide an ideal context for student exposure to interprofessional clinical practice and an experience of its importance.' [47], p2.
Microstudent and health service drivers There was only one example in the literature where the driver for the placement was to attract more students to undertake rural placements [39]. The literature did not detail any drivers for the provision of RCP from a clinical educator/supervisor perspective. For students and supervisors, the literature more frequently assessed the impact of a rural placement. These drivers were aligned to three distinct models of RCP identified in the evidence: 1. Placements designed to expose students to rural practice, rural health issues and rural lifestyle, and provide training in rural clinical skills (n = 5 [33][34][35][36][37];) 2. Placements designed to address community needs or fill gaps in service provision in rural and remote areas (n = 8 [18,[38][39][40][41][42][43][44],) 3. Placements designed to provide students with a specific skill set (n = 5 [35,[45][46][47][48][49],) Placements to expose students to rural practice, rural health issues and rural lifestyle, and provide training in rural clinical skills The key attributes of these models are summarised in Additional file 2. Placements designed to expose students to rural practice, rural clinical skills, rural health issues and rural lifestyle ranged from 1-year voluntary experiences in the fourth year of study [33] to 1-week compulsory placements in the third year of study [34].
Placements to meet community needs or fill gaps in service provision in rural and remote areas Often termed, 'service learning' or 'role emerging' placements [18,[38][39][40][41][42][43][44], the placement for speech pathology students in Broken Hill, Australia is an example of a placement designed to meet community needs [50]. Studentrun clinics in primary schools around Broken Hill were developed as a placement option for final year students to address concerns raised by the community about the lack of paediatric speech pathology services in the region [50]. As described in the eight studies examining placements to meet community needs, often these types of placements send groups of students to non-traditional placement sites such as schools or aged care facilities. Supervision is often less intensive, delivered as group supervision and therefore peer learning is frequently utilised to drive learning from placement.
Placements that provide students with a specific skill set The placements providing students with a specific skill set all related to RCPs designed to expose students to IP practice and to improve IP skills among students [35,[45][46][47][48][49]. These placements varied in structure; however, they tended to offer both discipline specific supervision and specific IP opportunities in group situations.
In what types of contexts do allied health rural clinical placements take place?
AH RCPs take place in a variety of settings and are organised and structured in a number of different ways and are designed to meet some or all of the identified drivers (Additional file 2). In summary, the following contextual elements were identified: the duration of the placement; single or multiple students (or multiple disciplines); practice setting; supervision model; mode of supervision; externally supported/facilitated placement; learning purpose; learning approach; level of choice (compulsory or voluntary); and the year of study in which placement is undertaken. There was little consistency in contextual features within each model of RCP, with no studies providing information on all identified features. Placements designed to meet community needs had the most consistent features, with a trend for multiple students to be placed at one time [38, 40-42, 44, 50]. Similarly, placements designed to provide IP skills tended to involve students from multiple disciplines [35,[46][47][48][49].
What mechanisms are responsible for successful delivery of rural allied health clinical placements?
Fifteen different mechanisms relating to the delivery of RCPs (Table 4, Additional file 3) were identified: 1) support for students; 2) support and recognition for supervisors; 3) external funding or sponsor; 4) sustained funding; 5) regional coordination/infrastructure and support (e.g. UDRH); 6) coordination role between university and placement site; 7) stakeholder engagement, consultation and partnership; 8) needs/demand analysis; 9) support for university placement staff; 10) selection criteria/student traits; 11) resourcing; 12) support from registration bodies and/or professional associations; 13) evidence based approach to placement model; 14) regular program planning, evaluation and feedback; and 15) student autonomy. The most commonly reported mechanisms across all placements were support for students [33, 35, 39-43, [33, 35-37, 39-44, 46, 47, 49] identified student support as: Information booklets and maps for the locality; travel and/or accommodation costs paid for; daily student allowance provided; induction provided at the beginning of placement; orientation session and tutorials; discussion of learning objectives; discussion of key concepts of rural health practice; accurate communication about what clinical experience in rural practice will offer to students; duration; pre placement reflection of personal strengths/weaknesses; post placement debrief opportunity; 1:1 supervision at the end of every clinical session; clusters of students being placed together; for IP placements, understanding of professional identity prior to placement; internet access; phone coverage; access to a library; provision of social opportunities; 2 Support and recognition for supervisors Provision of supervisor courses for local clinicians; providing support to supervisors during clinical placements; and provision of tutorial programs for students run by the UDRH/Rural Clinical Schools or universities. One paper identifies ongoing difficulties with health staff recruitment and retention impacting on capacity to provide consistent support for student supervision, particularly in rural areas where departments are relatively small [37]. Wolfgang et al. [37] describe how the UDRH provides support and education to clinical supervisors.  [41] identify that capacity to implement and evaluate the impact of sustainable RCPs that require collaboration between multiple stakeholders is keenly affected by whether or not funding sources are sustained [41,42].

Regional coordination / infrastructure and support
For example the Australian University Departments of Rural Health (UDRH) function as a single coordination point for the whole region and all the health organisations -'a one-stop shop for student placements' that involves streamlining administrative procedures, maintaining links with service partners, clinical supervisors, feeder universities and students [36,37,40,41].

Coordination role between university and placement site
Several papers [36,37,41,47] emphasized the importance of a central broker, advocate or 'go-between' in the success of implementing 'collaborative fieldwork' models that can increase the capacity of a clinical educator to take multiple students at one time. For example Wolfgang et al. [37] describe how a unit coordinator was responsible for meeting regularly with students or telephoned those in remote locations, coordinated the placement with the university and field work site, provided support and training to supervisors and organised accommodation and transport .

Stakeholder engagement, consultation and partnership
The importance of 'building meaningful partnerships' and 'monitoring that all roles and visions are clear and understood' were essential components of engagement with stakeholders when devising and delivering rural IP clinical placements. Kirby et al. [43] describe trust as a key factor for success: whereby high levels of trust was facilitated by close relationships between stakeholders which in turn was facilitated by social connection in the local community. Kirby states that the combination of work and social connection enriched levels of interaction and facilitated partnerships. The enabled a commitment to be investing and sharing resources. A Memorandum of Understanding to meet unmet need underpinned the partnership.
8 Needs / demand analysis As identified by Allan's study [38] describing university clinics, where needs analyses are not conducted there is a risk that the clinic may not provide a sufficient amount or range of clients due to poor geographic positioning of clinic within the campus, sporadic and ineffective marketing and/or lack of range of clients/problem types.
9 Support for university placement staff Two studies [38,41] describe the need to adequately support academic staff who run university clinics as their role is often stretched to cover both clinic operations as well as an academic load. Key to successful running of university clinics is also year-round running of the clinic, making support for aca demic staff even more pertinent to their success.
10 Selection criteria / student traits Moosa and Shurr [44] describe a placement opportunity for students to develop speech pathology resources in extremely remote and under resourced communities in Canada. They iterated the importance of a selection process to ensure the students had the aptitude and character to cope with the demands of the placement and the 'hands-off' supervision model utilised. The authors stated the following requirements for students to undertake the placement: "Interest in rural issues/working rurally, strong academic record, clinical placement evaluations that identified strong professional and ethical conduct, exceptional interpersonal communication skills, rapid integration of feedback, independent problem solving, critical thinking skills" (p.162). Interest in rural practice was also cited as a selection criterion [44].

Resourcing
Adequate resourcing for RCPs refers to the infrastructure, time, resources and staffing required to plan, develop, coordinate and deliver the placement such as: providing the placement venue (e.g. school/ health service), keeping track of and coordinating all student placements within the health service/ community setting and organising and delivering structured education and supervision opportunities (e.g. integrated clinical debrief sessions; group interprofessional sessions; case studies; online activities; [46][47][48][49] and stakeholder engagement, consultation, and partnership [33,38,40,42,43,[45][46][47]50]. The least common were student autonomy [40,44], sustained funding [41,43], and support for university placement staff [38,43]. These mechanisms have been mapped against the three different models of placement and are described below (Table 4 and Additional file 3).

Placements designed to increase student exposure to rural practice
The most common mechanisms reported for placements designed around exposure to rural practice were support for students [33,[35][36][37], regional coordination/infrastructure and support (e.g. presence of a UDRH) [33,35,37], and support and recognition for supervisors [33,37]. Support for students included provision of information booklets and maps for the locality; travel and/or accommodation costs paid for; daily student allowance provided; induction provided at the beginning of placement; orientation session and tutorials and more (Table  4).

Placements designed to address community needs
Community focused placements had a much stronger emphasis on community needs. Therefore, a needs/demand analysis [38,40,41,50] and stakeholder engagement, consultation, and partnership (in particular, development of 'communityacademic partnerships') [18,38,40,[42][43][44] were key to delivering these placements. For example, engagement with the local community in Broken Hill, Australia identified a need for paediatric speech pathology to help improve educational outcomes in children [50]. Sustained funding was also identified as a key mechanism for success and sustainability for the Capricornica chronic care placement model but was rarely realised [41].

Placements designed to provide students with a specific skill set
Placements designed around IP skill acquisition described the combined need for both support for students and supervisors [47,48], stakeholder consultation [45][46][47], and the designation of a coordinator role that liaises with both the university, placement site and other stakeholders [35,47,48]. IP acquisition placements are reported to be particularly resource intense. As such, further key mechanisms for successfully delivering these placements include the availability of funding and support from a funded agency (e.g. rural clinical school) to ensure adequate resourcing for planning, implementation and supporting students [35,47].
What measures have been used to capture the impact or effectiveness of different models of clinical placements, and what is the strength and quality of this evidence?
The 'success' of different placement models was measured in a variety of ways (Table 5 and Additional file 1) and included measurement of: educational and learning outcomes; student outcomes (such as satisfaction with organisation of the placement, the accommodation, information and support provided, and overall enjoyment); rural outcomes (such as intention to work in a rural area, employment in a rural area post study, knowledge and understanding of the rural context, attitude to living and working in a rural area); program outcomes (e.g. satisfaction with accommodation, support, pre-post placement expectations); supervisor outcomes; service and community outcomes (e.g. reduced waiting lists for Table 4 Mechanisms for delivery of different placement models (Continued) No Mechanism Description and journal clubs). Ongoing resourcing was linked closely to ongoing external funding, which was particularly important for placements designed to address community needs [34].
12 Support from registration bodies and/or professional associations One author cites that Interprofessional competencies need to be part of placement requirement/ university requirement as expressed by one participant "clinical training requirements are set by the universities who set requirements for placements-they don't require cross discipline work, so the hospital won't provide it" [43].
13 Evidence based approach to placement model Frakes et al. [41,42] describe using reviews of evidence and inviting international experts to present evidence and collaborate in research evaluating placement effectiveness in an effort to maintain quality placements.
14 Regular program planning, evaluation and feedback Regular evaluation against needs assessment is key to sustainability and success of placement, in particular for placements designed to meet unmet community need. Drawing from implementation science literature, Frakes et al. for example describe the need for a focus on evaluating all aspects of a new model (context, processes and interactions and capacity to sustain). The Capricornica model therefore uses of multi-level evaluation and feedback loops as mechanisms to monitor sustainability and success by collecting impact data around student, staff, patient, referrer and health service outcomes [41,42] 15 Student autonomy Student autonomy over determining community needs (conducting needs analyses) or developing the services and resources requested by the communities was key to student learning outcomes, particularly for placements designed to meet community need [40,44].

Placements designed to expose students to rural practice
These papers commonly reported a variety of student and/or rural outcomes. The evidence from these studies was generally of low quality with mixed, inconsistent results (Table 5 and Additional file 1). One good quality qualitative paper described students' improved understanding of the rural health context, improved professional skills and greater understanding of the role of other health professionals post rural placement [35], and another uncontrolled before-after paper reported significantly greater interest in rural work post rural placement for both rural and metropolitan origin students [34]. Brown et al. [33] were the only authors to describe the impact of a 12-month immersion experience for dietetic students on employment in rural settings post placement, with 50-100% of graduated students working in rural areas. The quality of this paper however was low.

Placements designed to address community needs
Community needs placements were measured in a variety of ways with an equal emphasis on exploring student and learning outcomes and the impact the placement had on the community. The evidence from all of these studies, although generally poor in quality, suggests these types of placements can have a positive impact on addressing community needs [38][39][40][41][42][43][44]50]. For example, the student speech pathology clinic established as part of the Broken Hill, Australia UDRH recorded that 231 primary school aged children were assessed in 2010; 58% of kindergarten children received a speech pathology intervention; and the number of new referrals on the speech pathology service waiting list decreased from 250 clients in September 2009 to eight in September 2010 .
Placements that provide students with a specific skill set (interprofessional) A variety of measures were employed to gauge the success of IP placements with IP educational and learning outcomes featuring consistently [35,[45][46][47][48][49]. IP outcome measures included: student and supervisor perceptions of IP learning outcomes (including Kirkpatrick's educational outcomes framework) [35,45,47,49]; and the IP education scale measuring student IP attitudes pre and post placement [45,48,49]. Most studies described successful IP outcomes for students as a result of their IP placement model situated in a rural area [50].
One study that used the IP education perception scale pre and post placement reported participation in an IP program in a rural community improved student IP scores; increased their understanding of others' roles; influenced attitudes towards IP practice for students and supervisors; and there was a significant increase in participants' positive perceptions regarding IP practice after they participated in the project [49]. As identified by McNair et al. [48], the context of the IP placement, described as the intensity of the 'immersion' experience, with students having to negotiate an unfamiliar environment to work and live together, also had a significant influence on students' learning outcomes. Uniquely, Gum et al. [47] identified that rural IPE placements also have a significant influence on student IP interactions with the rural community.

Strength and quality of the evidence
The majority of studies used a mix of post placement unvalidated self-report questionnaires, student placement activity measures and/or through interviews and focus groups with a variety of participants. Given the variation in outcomes measured and generally poor quality of the evidence (Additional file 1), robust conclusions cannot be drawn regarding the impact of different models of clinical placement, the exception being some of the IP placement studies that utilised validated IP outcome measurement tools [49] and good quality qualitative research designs [47] to explore the impact of the placement on educational IP outcomes.

Discussion
To overcome known limitations in drawing conclusions from a weak evidence base with significant variation in outcome measurement, this research aimed to capture the complexity of the context in which AH RCPs are operationalised by presenting the evidence using a logic model framework.
The logic model is presented in Table 6 and addresses the final question of the review: What is the relationship between drivers, contexts, mechanisms and outcomes? The logic model highlights the key ingredients that the evidence has identified as desirable for devising, implementing and evaluating a 'successful' AH RCP.
When connecting drivers with contexts and mechanisms to outcomes ( Fig. 2 and Additional file 1), there were only two clear identifiable patterns from the evidence. The first is when the driver for the RCP is to provide students with particular skills and competencies, IP placements undertaken in rural environments have demonstrated improvements in IP competence and increases in the number of student placement opportunities [35,[45][46][47][48][49] (Additional file 1). Key mechanisms that contribute to these outcomes include: the combined need for student and supervisor support, stakeholder consultation and engagement, and provision of adequate and ongoing resources and funding. The contexts or RCP features that support these mechanisms include: multiple disciplines and multiple students being placed on the RCP; an RCP that is 2 weeks or greater; and the RCP has a learning purpose specific to interprofessional skills. The strength of this relationship is supported by moderately good quality research evidence (Additional file 1).
The second identifiable pattern is when the driver for the RCP is to respond to community priorities of unmet health needs, such placements have demonstrated increased service delivery to 'in-need' communities and increased placement capacity [38][39][40][41][42][43][44]50]. Key mechanisms that contribute to these outcomes include: development of 'community-academic partnerships' alongside a community needs analysis. The contexts or RCP features that support these mechanisms include multiple students being placed on the RCP and the RCP having a learning purpose specific to exposure to rural contexts and rural practice (Additional file 1). The strength of this relationship is supported by moderately good to low quality research evidence (Additional file 1). Whilst not strongly represented in the evidence reviewed, these placements can be costly to provide and sustain [42,50] and, where communities are very underdeveloped and poorly resourced, student-led interventions were often not realistic or achievable [44]. Therefore, appropriate resourcing should be considered a key mechanism.
These identified patterns are complimented by other rural health workforce research. For example, eight factors have been identified that facilitate the development of effective and sustainable community-academic partnerships. These are (1) creation and nurturing of trust (2) respect for a community's knowledge (3) community defined and prioritised needs and goals (4) mutual division of roles and responsibilities (5) continuous flexibility, compromise and feedback (6) strengthening of community capacity (7) joint and equitable allocation of resources, and (8) sustainability and community ownership [51]. More recently, the following features have been identified as supporting successful and sustainable community health partnerships in rural and remote Australian locations: 1) identifying and responding to community need, 2) providing services of value, 3) community leadership and innovation, 4) reputation and trust, 5) consistency, and 6) knowledge sharing and program adaptation [25]. These elements should be considered in the development of any RCP concerned with meeting community needs. In cases when the primary driver for the RCP is to encourage students to work in a rural setting, there was no particular model or, indeed, any consistent contextual components that could be directly linked to increased rates of intention to work in a rural area. While many of the included papers listed attracting students to work in rural settings as being a key driver, only three papers directly measured it [33,34,37]. In these papers, there was no consistency of placement features (context) that could be linked to a stated intention to work in a rural area (Additional file 1). Where exposure to rural practice was an identified driver for RCP, this was measured in terms of attitude to living and working in a rural area and/or rural work readiness [33][34][35][36][37]. In these papers, providing good student support was identified as being necessary for 'successful' exposure to rural practice (Fig. 2, and see for example Paterson et al. [33] in Additional file 1). This association may lend some weight to the evidence that positive student placement experiences can play a key role in influencing the rate of rural employment of newly graduated nursing and AH practitioners [52]. Noting however that the strength of evidence to support these patterns is limited by the generally low quality of the research (Additional file 1).
There is also growing evidence that immersive RCPs (with multiple students placed for longer placement periods) can influence intention to work rurally [9]. However, other research indicates the decision of health professionals to work in a rural location is not determined simply by background or participation in 'excellent' rural placements, but varies between individuals as a result of the complex interaction of many factors [9,53]. For example one longitudinal Australian study shows that intention to work rurally increases over time, since graduation [54].
We have identified a number of drivers, contextual elements, mechanisms and outcomes, but there are significant gaps. The absence of some of these descriptors from our analysis may be because the articles simply did not include this information or did not see the need to include this information. There is opportunity, therefore, to use the context descriptors identified in this research to inform future reporting of RCP evidence.

Study limitations
This review has focused on placement models and interventions for rural and remote allied health practitioners and, as such, the mechanisms identified are limited to this group. Future research could consider a realistic evaluation approach that would integrate research examining rural recruitment strategies that include a greater range of health care practitioners and also successful rural recruitment strategies for workers outside the health care industry. The review focused on published, peer reviewed evidence and did not include evaluations or grey literature. Further, articles were excluded where no abstract was available to review. More extensive searching of the evidence base and grey literature may offer greater contextual richness to the logic model described here and allow for development and testing of propositions arising between the contexts, mechanisms and outcomes identified here [55]. The majority of papers identified in this review evaluate novel RCP models. Thus, there is an inherent bias in the conclusions that can be drawn from this review as there are few papers that describe more simple, 'bread and butter' RCPs, such as placements offered to just one student, hospital only placements, discipline specific placements or education outcome placements (e.g. paediatric placement in a rural setting). The terms "service learning" and "work integrated learning" which have been more frequently used in recent years in the Australian context to describe clinical placements, were not used in the search strategy. Whilst this review identified a number of studies that used these phrases, the findings of this study may be limited by this omission. Finally, the overall quality of evidence is low, limiting the impact of the study's findings. Whilst data were extracted and synthesised using program logic to overcome the lack of quality, there remains a critical need to invest in producing high quality research in rural contexts [7,56].
Future research should focus on testing the logic components identified in this review and developing robust proposition statements that can inform improved decision making around the contexts and mechanisms that contribute towards successful AH RCPs.

Conclusions
Whilst this review found some evidence to support the proposition that undertaking an RCP may lead to increased intention to practice in a rural area, there is little evidence regarding the type or model or elements of a RCP that can be applied to achieve this. Better quality research of AH RCP models is required. There is a need for more systematic and psychometrically robust measurements of the impact of different models of RCP. There is also a need to utilise more uniform, standardised and validated tools to measure key outcomes of RCPs, such as intention to practice in a rural location, rural work readiness, attitudes towards rural practice, and placement quality. Furthermore, defining, monitoring and consistently measuring sustainability as an outcome of RCPs is required. Finally, improvement in describing placements in a more systematic way to support comparison is necessary. The logic model presented in this paper provides such descriptors.