Paper selection (systematic review)
From the systematic search, a total of 2769 titles and abstracts were screened, which were filtered down to 117 full texts. The Google Scholar searches conducted on 11.03.20 produced 426,000 results on this specific day, and the first 40 pages of results were screened, resulting in four further papers. At this stage, 52 papers were included (figure 1).These were then screened for relevancy, i.e. whether these papers included sufficient descriptive depth regarding contextual factors, mechanisms, and outcomes underlying inter-organisational collaboration, which resulted in 35 included papers. Reference scanning and citation tracking resulted in a further four papers, giving a total of 39 papers included in this final analysis (figure 1) (1,4,12,13,15,17,22,34–65). Agreement between independent reviewers was 100%. Fourteen purposively identified papers were also drawn upon, which outlined MRTs, used to elucidate the workings of mechanisms, bringing the total number of papers included to 53.
Included studies comprised case studies (n=18), reviews (n=16), case control study (n=2), survey (n=2), and theoretical papers (n=1). Unfortunately, we were not able to identify any opinion pieces or commentaries that had sufficient analytical depth to inform the review. In terms of types of collaborations, covered mixed collaborative types (n=16), mergers (n=9), alliances (n=3), joint working (n=2), contracting (n=1), joint commissioning (n=1), integrated care (n=1), vanguard arrangement (n=1), accountable care organisations (n=1), community health partnerships (n=1), buddying (n=1), primary care partnerships (n=1), and combined trusts (n=1) (table 1).
Theoretical papers included one paper for partnership synergy (66), one for trust (67), two for conflict (68,69), one for power (70), one for coordination (71), one for leadership (72), two for organisational flexibility (70,73), one for task complexity (74), two for confidence and formalisation (75,76), and two for proximity theory (77,78).
Middle-range Theory and mechanisms
Frequently mentioned in seven of the systematically-reviewed studies was the concept of partnership synergy (4,15,36,37,39,53,63,65), which was first coined by Lasker, Weiss, and Miller (2001) as a means for explaining how partnerships achieve advantage over independent, competitive working. As such, this theory was adopted as an MRT, which explains how there are ‘partnership functioning’ mechanisms essential to explaining the processes of working together, as well as ‘partnership performance’ mechanisms which underpin the improvements that collaborating seek to attain. Lasker, Weiss, and Miller (2001) put forward partnership synergy as an intermediate outcome that comes after the functioning of the partnership but precedes the effectiveness of it (figure 2) (66). This means that when working well together, a combination of resources and skills of the partners is what enables achievement above and beyond what would have been possible individually. Partnership synergy can be considered a mechanism whereby a context of high partnership functioning leads to greater partnership synergy, and thus improved partnership performance. Improved partnership performance is likely to be an outcome in itself which results from mechanisms involved with an improved ability to achieve healthcare-related outcomes, such as reduced duplication of effort, economies of scale, and competitive advantage (66). But, these performance-related mechanisms will depend on the aims and structure of each individual collaboration. We also add to the MRT the concept of collaborative inertia, which was put forward by Huxham (2003) - one of the systematically identified studies. Collaborative inertia occurs when organisations and actors get ‘bogged down’ in the day to day functioning of the partnership (15). While trying to optimise the daily functioning, achievement of the actual aims of the collaboration fall by the wayside as significant manpower and time is devoted to partnership functioning rather than accomplishment of outcomes. It is possible that a collaboration will engage in a period of inertia in its earlier stages of formation, before synergy is later achieved. This concept of inertia was also put forward by a number of the included studies (15,17,52,53,63), and is defined as when “the output from collaborative arrangements often appears to be negligible or the rate of output to be extremely slow”(13; p.403).
This MRT theory, taken together, thus proposes that the mechanisms comprising ‘partnership functioning’ need to have their context configured very favorably before synergy, and thus, enhanced performance, can be achieved. As partnership functioning relies on many other contextual factors and the mechanisms that enable collaboration, these will be explored as elements that enable synergy and thus partnership effectiveness and accomplishment of its aims, in the following section.
Mechanisms underlying ‘partnership functioning’
The results of our analysis of the included studies identified a range of mechanisms underlying collaboration functioning, namely: conflict, trust, power, faith, interpersonal communication, leadership styles, cultural integration, and task complexity. We also identified a role for a ‘confidence’ in contract in certain circumstances. These mechanisms underline the ability for collaborations to perform through ‘partnership synergy’, and with avoidance of ‘collaborative inertia’. Although mechanisms can relate to both changes in reasoning and resources that an intervention introduces, the majority of mechanisms we identified relate to processes of reasoning by actors. This may be explained by our underlying assumption that ‘collaboration’ as an intervention in the inter-organizational setting is characterized by a change in organizational behavior from competitive to collaborative behaviors (79). The review also identified a range of contextual factors that affect how these mechanisms are activated (additional file 1). The following sections present the interactions between these various elements.
Building and maintaining trust was a key mechanism identified by 16 papers in the review, and trust can be affected by a number of contexts (4,12,13,15,36,37,47,52,53,57,61,80,81). As trust was mentioned so frequently, we sought to include an appropriate MRT for this element that explained how trust is linked to collaborative behaviour. Suitably, due to its use in many of the included studies (e.g. Axelsson & Axelsson, 2006), we identified Vangen and Huxham (2003) and their trust-building loop in inter-organisational collaborations as a suitable framework (31). Trust has been defined in a myriad of heterogeneous ways, but we draw on the concept as a key component of social structures (organisations), with trust being formed as a result of networks and norms between actors in the social structure (82). The trust-building loop proposes that a certain degree of trust is required to set the risk tolerance of each partner with respect to how ambitious the aims they agree to are. As more is accomplished by the collaboration, trust will be reinforced – but if failures occur, trust will be reduced; these successes or failures will again affect the risk tolerance in a cyclical manner (67). This concept of risk tolerance allowed us to understand how greater trust enabled a greater tolerance for riskier endeavours, thereby changing bit by bit to what degree a partner would be willing to act collaboratively.
Trust underpins the majority of decision-making that is undertaken in a collaboration, and also is tied keenly into other mechanisms such as respect, conflict, and power, which all may affect trust as an outcome (67). For example, as previously mentioned, every time a conflict occurs between organisations, it is likely that trust between them will be reduced (69). Trust is put forward by Vangen and Huxham (2003) to mean “the ability to form expectations about aims and partners’ future behaviours in relation to those aims” (67). Scholars argue that trust and risk are keenly interlinked, and trust is required to ‘take a risk’ in believing that a partner will do what is against their own interest for the collective good (83). This places the ‘trustor’ in a vulnerable position relative to the trustee – and in most voluntary collaborations this goes both ways. When results of these risks arise, they can build more trust or have it broken down depending on the outcome. A number of contexts are important for modulating the initial level of trust with which partners enter an arrangement as well – which can act as a buffer against future conflicts and task failures.
Trust building, synergy, and perception of progress
Trust building is another factor that needs to take place throughout the process of collaboration and is likely to be cyclical in nature, as acts which beget trust are usually reciprocated (67). Mutual successes such as achievement of outcomes reinforce trust in both parties (figure 3) (67,84). This loop was explicitly mentioned by included systematically identified studies (13,15,53). This means that outcomes need to be realistic and agreed upon by both parties, thus, if outcomes are too overambitious then trust will also be reduced as they are unachievable (figure 3). Likewise, this links into the mechanism ‘perception of progress’, which is defined here as how well organisational actors perceive the organisation to be progressing towards the aims of the collaboration itself. Perception of progress as a mechanism links into both trust and faith as outcomes and is affected by a number of contexts outlined below.
This CMOC was supported by quotes such as the following by Round et al. (2018; p. 300): “Challenges included a feeling that the programme had, “massively overambitious proposals in the original business case” and was “too ambitious with a lack of realism”. This hampered progress to deliver the initial objectives…” (46) as well as by Dickinson & Glasby (2010; p. 819): “the tendency to see partnership working as a panacea to a series of current problems, placing too much faith in its ability to deliver a series of over-ambitious aspirations, therefore running the risk of disillusioning staff if such aspirations are not achieved; and undermining the subsequent partnership by failing to attend to practical details” (1). Relatively unambitious intermediate aims and outcomes formulated at the beginning of a collaboration may thus serve to solidify and build trust early on, enabling achievement of higher ambition ultimate outcomes such as an improvement in care quality (85). This means that:
Unambitious aims (context) à better perception of progress (mechanism) à increased trust and risk appetite (outcome).
As such, in some cases:
Overambitious aims (context) à reduced perception of progress (mechanism) à conflict (outcome) (figure 3).
Trust is also essential to maximising collaborative synergy. As Jagosh et al. (2015) and Lucero et al. (2020) have identified in their analyses of public sector partnerships, without trust, partners will not be able to work together in a functional manner (86,87). This means that:
High trust (context) à partnership synergy (mechanism) à collaborative performance (outcome).
It is evident that a certain minimum level of trust needs to be maintained at all times for a collaboration to avoid dissolution, and that certain factors are likely to modulate the level of trust already in place when people begin to initially work together. As mentioned by papers in our systematic review, these factors could include whether the organisations involved have had pre-existing collaborations that were successful (or not) (13,44,52,56), as well as the historical context of collaborations in the geographical area in which the organisations are located, and a partner’s reputation (4). These factors have the potential to act as enablers or barriers to potential collaboration by modulating the pre-existing level of trust and suspicion with which partners will begin collaborating. This was supported by quotes such as the following from Auschra (2018; p. 7): “if they have gathered experiences from former collaborations, organisations assess cooperation outcomes differently” (4). Thus, the following CMOC emerges:
Existing successful collaborations (context) à better initial trust (mechanism) à greater ambition in objectives (outcome) (figure 3).
The degree to which a collaboration is formalised was mentioned in included studies as a method of instilling trust between partners by cementing tasks and accountability in contractual, legal terms based on relational contracting (44,53). Formal agreements forged at the beginning of such arrangements in the Connecting stage of collaboration can also serve as a scaffolding which holds up and solidifies trust between partners (figure 3) (84,88). This is because, as rules are laid down with a legal mandate to uphold them, there is an understanding that the other side will follow them, and is supported by quotes such as the following by Casey (2008; p. 78): “the more formalized a partnership is, the more likely it is to be maintained, because formal arrangements tend to signal commitment and accountability” (47). Thus, we hypothesise that:
Legal agreements (context) à greater initial trust (mechanism) à greater risk threshold and perception of progress (outcome).
A deeper look into the potential impact of formalisation on trust, depending on the collaboration type, is in the discussion.
A further mechanism explicitly mentioned throughout the included papers was conflict; many factors lead to conflict if not properly managed, including cultural differences, the management of individualist vs. collectivist interests, power dynamics, congruence of aims and objectives, whether collaborations are dissolved as appropriate, ongoing evaluation, organisational ownership of decision-making, and the pace of collaboration development (4,13,15,46,47,55). Ideally, all of these factors are overseen by conflict resolution mechanisms that rely upon mutually agreed governance and accountability arrangements between partners. However, there are also other mechanisms at play that can prevent conflicts arising before they even happen, for example, developing cultural integration plans that ensure that conflicts arising due to cultural differences in workforces are planned for and mitigated (60).
As mentioned by Lumineau, Eckerd, and Handley (2015), conflict between organisations is often very different from inter-personal conflicts due to the level of interaction, decision-making parties, incentives and motivations of key stakeholders, governance structures for preventing and managing conflict, repair mechanisms available for resolution of disputes, and the institutional context. Essentially, the situations become much more complex due to the myriad actors and mechanisms involved. Conflicts can also take numerous forms, such as whether they are competence-based (relating to skills or knowledge of partner) or more fundamental, integrity-based conflicts (69). These have differing implications for how resolvable they are with different management strategies.
Perhaps the most pertinent categories of management strategies include constructive (joint problem solving and persuasion) vs. destructive (domination) conflict resolution strategies which are clearly evocative of the type of relationship that is at play between partners (68). The outcomes of conflict are typically a function of the effectiveness of the conflict resolutions in place and the type of relationship which already existed (69). As already mentioned, this could manifest in a loss of trust, or in the case of re-commitment that arises from a constructive management process, could potentially lead to improved trust due to a gain in collaborative working spirit. We suggest that conflict is keenly linked to trust, and that conflict can be both a context and a mechanism depending on the element of analysis (figure 3). For example:
Included papers (e.g. Murray et al., (2018)) suggest that:
Conflict between partners (context) à can lead to reduced trust (mechanism) à reduced ambition and faith in the collaboration (outcome)
Likewise, others (e.g. Auschra (2018)) suggests that:
Having a shared vision (context) à is likely to reduce conflict (mechanism) à leading to improved trust (outcome)
Conflict (context) à Constructive conflict resolution strategy (mechanism) à lowered reduction to trust (outcome)
Conflict (context) à destructive conflict resolution strategy (mechanism) à reduction to trust (outcome)
Accountability & commitment
As discussed previously, accountability and conflict resolution mechanisms are key and should be established as a part of the governance of the arrangement in the Planning phase of a collaboration (84). Effective conflict resolution and accountability processes are essential to modulating the impact that conflict has on the collaboration itself, as mentioned by studies in the systematic review (47,53,65). As conflict causes loss of trust, loss of faith in the collaboration, and loss of perception of progress; but, if effective measures are in place, then the likelihood of conflict spiralling out of control and causing the downfall of the partnership is much lessened. This potential for dissolution and loss of trust is supported by quotes such as the following from Murray et al. (2018; p. 775): “For Access ACO (Accountable Care Organisation), these tensions were resolved through active conflict resolution, and the alliance remained intact throughout our research period. In contrast, for Collaborative ACO growing distrust paired with the management partner’s decreased investment and high fees prompted dissolution” (47). Likewise, the inverse is true as well (69). As such:
A context of preparedness and accountability (context) à reduced conflict (mechanism) à a smaller reduction in trust (outcome) (figure 3).
The literature reported that conflicts can also arise within organisations involved in a collaboration, which have the potential to reduce the organisation’s effectiveness (12,54,59). For example, workforce churn brought about by people leaving the organisation, due to the additional workload brought about by the partnership, or due to other factors such as pay imbalances, will likely lead to conflict within an organisation and reduce organisational effectiveness (53,61). Likewise, if lower-level staff are not involved in the decision-making around collaborative involvement and the shape the entity should take (context), then there may be conflict (mechanism), which could lead to a reduction in faith (outcome) (12). Rather than in the case of inter-organisational conflicts, if managed prior to getting out of control, conflicts within organisations can likely be dealt with without involving the other partner. Factors such as workforce churn could easily be noticed by another partner and could lead to conflict in that case, however, if managed properly, should not escalate to such a degree. Intra-organisational conflict reduces the ability for an organisation to accomplish the aims of a partnership by wasting organisational time on conflict resolution and by reducing faith in the partnership. In this case:
Intra-organisational conflict (context) à reduced perception of progress(mechanism) à reduced faith (outcome).
Power was mentioned throughout the included papers as a key mechanism underlying collaborative efforts (4,13,17,47,61). Power refers generally to the influence one organisation has over another, and can stem from hierarchical position, control over critical or scarce resources, and from discursive legitimacy, or ability to mobilise external support (89). Power relations are also key to trust building; some arrangements can be characterised by a dominant partner controlling the agenda to protect its own interests. This is supported by quotes such as the following by Murray et al. (2018; p. 767): “A power-sharing approach and consistent investment in the community with support for local ACO-level decision making fostered trust at the leadership level between the ACO and the management partner” (47), In lopsided relationships in terms of organisational size, the larger one may dominate (67). This has the potential to skew the trust relationship by lowering the initial degree of trust. In cases where collaboration is enforced by a governmental organisation, such as with buddying or competition-related acquisitions in the UK’s NHS, these power dynamics may be intrinsic to the relationship and therefore it may be very difficult to build trust (8,90). Willem and Lucidarme (2014), in their review and test of the role of trust in inter-organisational networks, propose that mandatory networks are likely to be less effective and have reduced levels of trust.
Lopsided power in relationship (context) à domination by one partner (mechanism) à reduced trust (outcome)
Reduced trust (context) à reduced risk threshold and aim ambition (mechanism) à reduced achievement (outcome)
Douglas (1998) posits that resource exchange during partnerships relies keenly on the power dynamics within the relationship. A more dominant partner may take more resources for themselves, or dependencies may develop whereby the ‘weaker’ partner is dependent upon the stronger one for resource, which could be called an ‘unhealthy’ power dynamic (47). This ‘unhealthy’ power dynamic in which one partner dominates is characterised by any scenario in which another partner has a loss of trust arising from the dynamic. In one of the cases of a healthcare alliance analysed by Murray, D’Aunno, and Lewis (2018), cost savings that were garnered by the alliance were sequestered by the management partner who was dominating in the power structure, fatally reducing trust and causing the end of the alliance (47).
Unequal resource distribution à domination in power hierarchy (mechanism) à reduced trust
Related to trust is the concept of faith, which may also be expressed as confidence or belief in the collaboration itself. While trust always relates to inter-organisational relations and belief in one’s partner, faith relates more to how actors within one or more organisations continue to believe in the collaborative endeavour as something of value. While trust is likely to modulate faith to a certain degree, as low trust in a partner could affect faith in the partnership, a scenario is also foreseeable where a partner has low trust but high faith. One partner could have been let down repeatedly by another partner in the achievement of the aims of the partnership, leading to low trust, but is not yet ready to give up on the concept of collaboration itself, as the plan is strong and the logic for how collaboration can achieve the intended outcomes for stakeholders is still clear. The concept of faith was not explicitly articulated in any of the included studies, but some have highlighted the importance of confidence and belief in the collaborative arrangement, which is a roughly analogous concept (37,47,48). We posit that faith/confidence/belief is a distinct mechanism from trust that , based on included studies, is chiefly affected by two contextual factors: ambition, and authenticity of the collaboration. These can serve to modulate the faith which actors hold in the collaboration. It is possible that there is a faith-building loop that exists within each involved organisation in a collaboration, similarly to the trust-building loop. In this sense, faith is also essential to collaborative synergy as it upholds collective desire to work on collective goals. As such:
High faith (context) à partnership synergy (mechanism) à collaborative performance (outcome)
Low authenticity of the collaboration (context) à low faith in arrangement (mechanism) à lower partnership synergy (outcome)
Ambition and authenticity
We suggest that Ambition, , is to what degree the aims and outcomes set in the planning phase of the collaboration are realistic (and feasible) (1,37,46,52). The degree of ambition needs to be kept realistic to ensure that there is a perception of progress and the building of trust between partners, as well as to maintain faith in the relationship. As Round, Ashworth, Crilly, Ferlie, and Wolfe (2018, 300) mention in their case study of an integrated care programme, the initial plan was “too ambitious with a lack of realism”, and this “hampered progress to deliver the initial objectives”.
Ambition is too high (context) à reduced aim achievement (mechanism) à reduced faith and trust (outcome)
Authenticity is another consideration, and it refers to whether the collaboration actually is based upon a real need to solve a problem, or whether a collaboration is simply undertaken to look ‘trendy’ and virtuous (1,61). Inauthentic collaborations are unlikely to inspire workers to put significant effort into collaborating.
Inauthentic collaboration (context) à reduced faith in collaboration (mechanism) à reduced aim achievement (outcome)
Increased coordination is often one of the primary motivations for organisations seeking to cooperate. Coordination refers to a reduction in duplication of effort, reduction in gaps of services, and sharing of knowledge and skills (4,52,71). The degree to which organisations are coordinated is a key mechanism underlying success and failure of collaborations. Coordination is a mechanism that is linked to mechanisms of information exchange and interpersonal communication.
A common concern in the literature is the ability to exchange information between partners as required, which is often key to properly coordinating delivery of care and other aspects of work (4,17,54,57,61,91). A lack of information exchange leads to a lack of coordination:
If information is not exchanged as required (context) à a lack of coordination can occur (mechanism) à leading to conflict (outcome) (4,54).
Key to the sharing of information is the interoperability of, and devotion of resources to, information technology systems (17).
If more interoperable systems are already in place (context), à a reduction in task complexity (mechanism) à will make combining these systems more straightforward, improving perception of progress (outcome).
Dialogue between actors is required in order to build the trust required by the collaboration, transfer information, and properly coordinate tasks (4,53,61). However, just as communication is required to build up trust, so too is trust required for actors to desire to communicate (53). The essential role of communication in building trust is supported by excerpts such as the following by Wildridge et al. (2004; p.7): “The role of clear, consistent communication (in trust-building) is at least implicit and sometimes explicit in much of the literature” (13). As such:
Greater interpersonal communication (context) à increased trust (mechanism) à increased synergy (outcome)
However, this element can go both ways and depends upon the culture of the actors interacting.
Conflicting cultures between partners (context) à then increased interpersonal communication (mechanism) à may lead to conflict (outcome)
Likewise, the inverse is also possible (i.e. with more related cultures leading to reduced conflict).
Interpersonal communication is also essential for collaborative synergy to be achieved and, thus, for performance to be maximised. As Lasker, Weiss, and Miller (2001; p194) put forward, “Effective communication strategies and mechanisms to coordinate partners’ activities are needed to facilitate synergistic thinking and action” (66) . Thus:
Increased interpersonal communication (context) à partnership synergy (mechanism) à increased collaborative performance
According to the included papers, communication with and involvement of stakeholders can make the difference between a collaboration being taken seriously or not (13,36,40,48). Inclusion of these perspectives allows for definition of the correct priorities and focusing on delivering benefits where they are needed most. We suggest that the ‘engagement of stakeholders’ context is keenly linked into authenticity of the partnership, increasing faith, congruence of aims/objectives, and focusing on the right outcomes. This is supported by excerpts such as the following from Wildridge et al. (2004; p.7): “Inclusion of service users’ perspectives, for instance, can make the difference between a project being taken seriously or not” (13). The following CMOC is a result:
Involvement of stakeholders (context) à increased authenticity of the collaboration (mechanism) à increased faith in the collaboration (outcome)
Eight of the included studies mentioned leadership as being key to the success of partnerships, however, few elucidated upon why (13,34,36,47,49,56,58,61). Evans and Killoran (2000), in their realistic evaluation of five different models of partnership working, mention “leadership and management skills” being enhanced through “external consultancy support” and “strong project leadership” but do not outline the mechanisms underlying these aspects. Wildridge et al (2004) mention “continuing, visible, and joint commitment from individuals in positions of leadership and influence” as very important, and Leach et al. (2019) mention in their evaluation of a buddying programme that “compassionate leadership” was key. Leadership is essential to any kind of organisation regardless of whether they are partnering with another but is not in all cases a mechanism that underlies partnership, rather, can perhaps also be a context that frames the partnership from the outset. Nonetheless, it is possible that collaboration can introduce a change in leadership style towards one that may maintain a collaborative endeavour better than others. In this case, it would become a mechanism in a realist sense. Due to the lack of description of mechanisms, to identify those relating to leadership, we turned to the theoretical literature.
According to Fillingham and Weir (2014) in their study of integrated care leadership in the UK, leadership during partnership requires different skills than those displayed during their climb of the organisational ladder. Successfully collaborating requires use of individual skills rather than their position, an ability to compete in a way that enhances the competition (through collaboration), conducting of business ethically in a way that builds trust, and development of a focus on process emphasising the ‘how’ rather than the ‘what’ (72). Likewise, Hunter and Perkins (2012) broadly agree, emphasising a participative and open leadership style of listening, asking questions, and co-producing solutions – a less dictatorial style (92). Huxham (2003) significantly expands on this by analysing the activities which leaders should be focusing on to move a collaboration forward, namely, facilitative behaviours which serve to involve and mobilise members. Huxham (2003) also identifies a more combative leadership approach which may occur in collaborations that have unequal power dynamics (such as those which are mandated), in which the leader engages in “collaborative thuggery” to push out those who do not align with their vision of collaboration. We posit that this, too, may work in cases where trust was compromised from the outset and collaboration is mandatory in the first place, but would be likely to undermine trust and respect in voluntary collaborations. Depending on the situation, leadership may act as either a context for another mechanism, or a mechanism in itself which allows collaboration to flourish or flounder (table 3).
Collaborative leadership style (listening, asking questions, co-producing solutions) (context) à improved trust (mechanism) à better aim achievement (outcome)
Mandated collaboration and very low trust (context) à combative leadership approach, pushing out those who do not agree (mechanism) à shared vision for partnership (outcome)
Voluntary collaboration (context) à combative leadership approach, dominating power hierarchy (mechanism) à reduced trust (outcome)
Additionally, others have drawn leadership keenly into the concept of developing and integrating culture, which is key to establishing greater trust and respect between workforces and leaders (93). Factors relating to leadership which affected successful creation of a new culture included: establishing organisation-wide communication channels and outlining outcomes for different staff types, involving more willing partners first, and leading in a positive and constructive manner (93).
Positive leadership style (context) à easier integration of cultures (mechanism) à greater trust (outcome)
While there are no doubt further, more detailed behaviours key to leadership styles that will uphold collaborations, these will be explored in greater depth in the next phase of theory refinement.
Culture, defined here as the attitudes and beliefs held by a workforce, is often cited as a primary reason for dissolution of collaborations if conflicts arising from differences in culture are not properly mitigated (4,17,37,47,51,54,81). Included studies suggest that auditing all organisations’ cultures, performing training around values and behaviours, implementing ongoing measurement, and even hiring and firing based on values may all contribute to general ‘preparedness’ to avoid conflict (47,53,65). Thus:
Preparedness for conflict through cultural similarities (context) à reduce conflict (mechanism) à reducing the impact on trust (outcome).
Likewise, the distance in culture (which could be measured if one desired) is also likely to modulate the ease of integration in this respect. Thus:
Cultural closeness (context) à reduce conflict (mechanism) à avoiding degradation in trust (outcome).
Of course, as trust gradually reduces it may reach a threshold at which dissolution of the collaboration occurs and collaborators recede to competitive behaviour.
According to Kopanaki and Smithson (2013, 95), organisational flexibility refers to an organisation’s capability to face environmental disturbances, or adapt when confronted with new circumstances. Of course, flexibility will be a trait inherent to the organisation before it even enters the collaboration, which will modulate the ability, and speed, with which an organisation can pivot to working together collaboratively. As such, we have considered it a context rather than a mechanism per say. Flexibility and/or capacity were mentioned frequently in the included studies of the systematic review (13,35,52,54,56,65,91). According to What Works Scotland (2015, 8), “one of the most striking themes emerging from analysis of this results chain is the need for effective partnerships to develop clear structures and processes whilst allowing for flexibility, engagement and responsiveness” (35). Willem and Lucidarme (2014) put forward the idea that collaborations are oftentimes intended to be more flexible alternatives to the status quo, and that low flexibility can lead to an overly bureaucratic process, reducing trust. We suggest that flexibility is not a mechanism, but rather, a contextual element that impacts how well collaboration can be implemented. However, flexibility may be able to be enhanced through other means, and those means may be mechanisms in themselves. Therefore, we posit that:
Greater organisational flexibility (context) à increased trust due to improved goal achievement (mechanism) à reduced conflict (outcome) (table 3)
Perception of task complexity
Perception of task complexity is a mechanism that underlines how actors perceive how difficult tasks are to complete, with more complex tasks requiring both more resources and more manpower to achieve (74). Complexity as a mechanism was referred to by papers such as Kendall et al. (2012), which refer to “diversity and complexity of the problem” as a key factor influencing success of collaborations, likewise, Mandell & Steelman refer to the ‘complexity of purpose’ underlying how difficult the aims of the collaboration are to achieve (44,65). This perception of difficulty is likely to feed into faith (i.e. the belief in the collaborative endeavour). Thus:
Perception of great task complexity (context) à lower faith (mechanism) à reduced synergy (outcome)
Organisational size is a key factor that the included studies proposed affects the initial complexity of the collaborative arrangement (38,94), as well as the size or type of the problem that collaboration is intended to solve (44). Gaynor et al. (2012) used econometric modelling to assess the characteristics and impact of 102 acute hospital mergers that took place in the NHS between 1997 and 2002. They found that compared to matched control hospitals, mergers tended to involve smaller hospitals with weaker financial performance. The main impact of mergers over the subsequent four years was a reduction in capacity and associated activity, with comparatively little impact on a range of performance measures, and there was little evidence of an effect of size on merger success. Fulop et al. (2002) also performed a qualitative study of mergers and found that increased size provided benefits in terms of having a larger pool of professional staff, increased attention from local authorities, and better cross-fertilisation of ideas. However, increased size also led to more remote senior managers, not enough cohesion through multiple levels of the workforce hierarchy, and a loss of informality and autonomy felt by those moving from smaller to bigger organisations. They found that:
Merger of larger organisations (context) à more task complexity (mechanism) and à slowed decision-making and internal communication (outcome).
As such, the organisational size is a context that primarily modifies the mechanism of task complexity, and as discussed above, achievement of more ambitious tasks requires greater risk threshold, meaning a requirement for greater trust (67).
Larger organisational size (context) à greater complexity of task (mechanism) à greater trust requirement between partners (outcome)
Ease of acquiring partner
On the whole, inter-organisational collaborations are entered into to solve a problem through some achievement that collaboration enables. Mandell & Steelman (2003) mentioned that collaborations founded upon solving simpler problems may make it easier to get potential partners to the table, as they may feel the stakes are low enough to ensure their individual goals can still be maintained and the expected difficulty of the task is reduced (44). Simpler problems may also link into trust, as easier achievement early on in the process can reciprocally foster more trust between partners (figure 3).
Simpler problem to solve (context) à reduced complexity of task (mechanism) à lesser trust requirement between partners (outcome)
Regulatory bodies, i.e. those above the collaborating ones in the hierarchy (i.e. governmental or health regulation authority bodies), can impose legal or resource constraints, or incentives, for collaboration to occur. This was cited frequently as both a barrier and enabler for collaboration (4,44,81). For example, in the UK, the Competition and Markets Authority has posed a barrier to circumvent for NHS providers seeking to collaborate by mandating a certain degree of competition to be in place (8). Auschra (2018) provides evidence that these barriers can manifest by forbidding collaboration entirely, stopping it before it even begins, or causing additional time and financial cost considerations (i.e. legal problems hampering information exchange, pooling of budgets, and bureaucracy requirements). For example, this is evidenced by reflections by Das-Thompson et al. (2020; p. 26) regarding moving towards system-level collaboration in England: “current regulation of individual providers is acting as a barrier to integration, with limited incentives to encourage wider performance implications at system level”. Therefore, it is possible for a:
Favourable regulatory environment (context) à reduce task complexity (mechanism) à enhancing perception of progress (outcome).
This will go on to improve trust and faith. Vice versa is also possible, i.e. that an unfavourable regulatory environment increases task complexity, due to its presence as a barrier and facilitator in the literature.
Geographical proximity of partners has been lauded as a contextual element which can lead to failure if not considered highly enough in the initial phases of a collaboration (4,42,54,81). Geographical co-location enhances casual face-to-face interactions (both planned and impromptu) and, can be responsible for “encouraging informal contact, which increased mutual understanding” (68; p.230). Geographical proximity therefore fosters information transfer and trust through improved mutual understanding (77,78). However, some outline that it may also lead to greater informality which has potential to undermine any perception of progress (54). As such:
Greater proximity (context) à enhanced interpersonal communication (mechanism) à fostering trust (outcome).
Point of entry into collaborating
The point of entry into a collaboration is highly likely to shape the nature of the relationship that follows (14,95). In addition to the contextual factors in additional file 1, our review sought to identify entry points and drivers for partnering in included studies and categorised them using bottom-up thematic analysis into 6 categories (table 2). In the case of seeking to partner due to funding and resource concerns, market opportunities, or innovation, a collaboration is likely to be shaped primarily by the regulatory environment, history of prior collaboration, and stereotypes from knowledge of other collaborations in the area. All of these factors together serve to form the context to determine the initial level of trust.
In the case of existing interdependence, it is likely that it will be a collaboration moving from a less integrated form to a more integrated form, i.e. they may have been members of a network but are now moving to a full merger. In this case it is unlikely to be considered unless they already have a positive relationship, thus, the initial level of trust is likely to be elevated. As mentioned previously, policy directives (i.e. mandated) partnering, are likely to drastically lower trust and power dynamics from the outset. This is supported by Auschra (2018; p. 7), who states “if a (mandated) collaboration threatens the political and economic interests of an organisation involved, it can be very reluctant to collaborate” (4). Likewise, entering a collaboration primarily with the goal of avoiding marketplace threats is also likely to lower trust, as an organisation will still be entering the arrangement from a standpoint of relative negative performance and fear.
The role of entry points and drivers will be fully explored in a future paper as another part of our realist synthesis, due to length considerations and the requirement for further systematic searching. However, in this current paper we posit that:
Mandated collaboration (context) à creates power imbalances (mechanism) à reduced initial trust (outcome).
Mandated collaboration is likely to require significant early ‘wins’ to enhance perception of progress, and mutual respect/goodwill to overcome. This will be explored further in the discussion.