Commissioning context: the challenges of network leadership
The current commissioning context presents a number of challenges for leaders in establishing innovation networks. In the following analysis, we examine the dynamics of multiple relationships which CCG leaders needed to establish to facilitate commissioning across their health networks, and in particular the need to enable knowledge exchange, network coherence and network stability as dynamic capabilities that support innovation networks. We consider the CCG board relationship with PCTs, health providers and service users, frontline GPs and the broader health polity. We conclude our analysis by comparing two innovative developments in CCG board commissioning practices in the sites studied, using them as illustrative rather than exemplars.
Establishing relationships with PCTs
There were variations in viewpoints among the CCG leaders at the six sites as to the way relationships with PCTs were managed. Leaders in some sites worked well with PCTs describing the relationship as cooperative and being ‘open’ and ‘supportive’, ‘getting better at seeing each other's point of view’. Some CCG leaders viewed PCT employees as a useful source of information and commissioning expertise. A CCG board member at site B pointed out: ‘I see my role as coordinating, having some ideas and then asking PCT people to develop those ideas. There’re only so many hours in a week and I can't do everything, so I draw on the skilled people at the PCT’. At site C whereby leaders developed a novel and collaborative arrangement whereby GPs and PCT managers were paired together to form a PCT sub-committee to resolve commissioning issues. These examples reveal the important role many PCT staff played as knowledge brokers who facilitated knowledge sharing and transfer across the network.
At the same time, however, GP leaders were acutely aware of the perceived limitations of the knowledge held by PCTs in commissioning. It was generally understood by CCG board members that PCTs were ‘being abolished [because they] haven't delivered what [they] should have done” (site B). A GP in site A was similarly critical pointing out that: “the contracting has been poor and it hasn't been adequately informed […] it is basically a legacy […] There wasn't actually any thinking or decision making’. Thus leaders were wary of adopting the knowledge and ideas of PCT commissioning practices.
A similar dilemma was faced by the PCT staff. On the one hand they recognised that ‘you've got the PCT trying to offload its activities to the CCGs’, in a supportive manner. On the other hand, a number of PCT employees felt threatened by CCG formation and were highly aware of their own job insecurity. As a result PCT members were not always willing to openly cooperate with CCG leaders, for example restricting funding of new commissioning arrangements. A GP described how the indifference of PCT employees towards the success of the CCG led to frustration in his board. There was a perceived view that a ‘not invented syndrome’ limited the potential for innovation; ‘nobody got the idea and they just refused to fund it’ (site C). The wavering support of PCTs stemming from the uncertainty of their future contributed to instability across the health network. There was also system-wide concern as to who would be responsible for the essential non-commissioning tasks currently being done by PCTs, and how they would be undertaken in the new health system. This hindered the development of trust and commitment as a critical basis for collaborative relationships with CCG board members.
Co-location arrangements further constrained (or enabled) communication between CCG members and PCT employees, leading to misinterpretations and delays in the transfer of information and data. One PCT director (site D) felt that their good relationship with GP leaders ‘was due to geography […] we brought the PBC support unit into the PCT building so they are in the same place as us […] sitting side-by-side with the PCT staff […Now with CCGs] that absolutely helped’. In our research sample, sites that had supportive relations between respective PCTs and CCGs used the PCT premises to hold their board meetings. In networks where CCGs were detached from PCTs, the board meetings were held elsewhere (eg, in sites B and E). These results are reflected by the social network analysis (see comparison between sites A and B in ).
Site A and site B network ties
Relating to providers and users
As discussed later on in our vignettes, a critical CCG leadership task is embedding relationships with health providers within the commissioning network, integrating secondary care provision with primary care in novel ways. A CCG board member suggested: “We need to get that relationship (commissioner-provider) off from a good start […] to sort a strategy that is going to pull them (providers) in from the beginning [and realise] that it's not ‘take all our money and continue to deliver as you've always done’. We've got to do things differently” (site B). Establishing trust and adequate knowledge exchange between CCG and provider entities remained an on-going challenge.
See online supplementary box 1 highlights multiple instances from our analysis regarding the importance of knowledge exchange and collaboration in enabling service integration across primary and secondary care structures.
Further, our results from the SNA analysis showed that there was generally a substantial lack of communication between CCGs and acute providers. In our CCG network sample, boards had a maximum of three ties with acute providers; this is very low when considering that these relationships are at the core of clinical commissioning and central to all the sample local innovations, including those summarised in the vignettes.
Another important network dynamic between CCG boards and healthcare providers related to knowledge sharing around appropriate level and type of costing data relevant to commissioning. This lack of information often described as ‘a black box’ around the services being provided and their associated costs leads to challenges of network level coherence of information to support innovation. A GP board member at site C explained, “we have actually no idea what the costs are of these pathways…it is very difficult to get any data or real information from [the acute provider]…they haven't had to share this before we can't commission [properly] without it.” Another GP board member reinforced that even in their own medical practice it was difficult to manage patients’ care in a way that optimised commissioning efficiency; ‘When I sign the referral letter I commission the spending of that money, but effectively what I'm doing is signing a blank cheque because I have no idea what the cost will be as the patient goes down that pathway. And if say there were two competing providers … which of those two pathways would be better to use and what are the costs and the outcomes of the two pathways, well I don't have that information’. As discussed later in the vignettes, comparative information and data analysis were important initial drivers of the innovation process. The tension between GP commissioners and secondary care specialists is described by a PCT employee as a conflict of interest where ‘providers want to maximise their income while [commissioners] want to maximise efficiency’ (site B).
In addition to providers, users constituted another important stakeholder that contributed knowledge towards the commissioning process. Over half of the CCGs had a patient representative on their board (sites A, B, E and F) to improve the final service offering. One of the patient representatives interviewed felt that he made ‘direct input’ into the board meetings, and felt that he made an important contribution as ‘any service user knows what it's like on the other side, to be on the receiving end. They can give very practical suggestions about what works, what doesn't, what are glitches in the system’ (patient representative, site A).
However, there was voiced confusion among leaders regarding how experiential knowledge from service users should be used and incorporated into the wider, population-level commissioning agenda of CCGs. A GP leader (site C) highlighted it was a challenge to engage patient groups into providing inputs at the locality and or CCG level: ‘Patients are not usually interested in it’, ‘they are busy and do not want to do things like this’ (site B), ‘patients will only be involved if there is money to be made’ (site A). In addition, many GPs commented that when inviting patients to provide feedback ‘you get half a dozen […] with particular reason or agenda’, suggesting this form of engagement did not lead to constructive dialogue on improving patient care. A GP from site B pointed out: ‘I think they [patients] are just there representing their own views as they see it’. Even though the wider perception from policy documents on public and patient involvement in commissioning was that patient views were valuable, there was no mechanism in place to operationalise lay representation and overall it was often carried out in a piecemeal fashion. For example, in some of the locality meetings we observed, individuals who had the flexibility to attend were listening attentively to discussions without engaging in overt dialogue. In other meetings, there was a set time given to patient representatives to present their perspectives. As such, several GP leaders felt that in the current fiscal climate and organisational upheaval, investing scarce resources in organising patient groups and their input was questionable, revealing the challenge in genuine public or patient representation.21
Engaging with frontline GPs
In order for commissioning decisions to reflect the corpus of primary care views across the network, CCG leaders need to find mechanisms for knowledge exchange with frontline clinicians. As shown though both case vignettes of innovations uncovered within our sample CCGs, novel ways of delivering a service or new services entailed commitment and engagement of frontline GPs, both in providing the new ideas and also enrolling colleagues in the new practice. Enabling knowledge flows across the network also enables the development of innovative ideas. Engagement and nurtured relationships with frontline GPs helps ensure that the knowledge held by these members is made available across the network, contributes to new practices and guides the leaders’ decision making.
Yet the ability to engage with front line GPs is related to the CCG size; smaller networks can more easily be densely connected, as it is easier to maintain ties with a smaller number of individuals. Network size in our context, is directly related to the proportion of population covered as government payments follow the patients. In the commissioning context, larger networks create a more stable environment (ie, network stability) as risk (in particular for financial failure) can be spread across the whole network. This more stable position also improves the leaders’ ability to negotiate, owing to their increased purchasing power across the network. “There's this sense that we have to be big in order to have the clout to negotiate” (site B). However, as network size increases, it becomes more difficult for leaders to engage with frontline members. Thus leaders also kept stressing that ‘if they [completely] ignore the size issue, they will fail to get [GPs] engaged and on board’ (site B), thus, highlighting the difficulty of engaging frontline GPs in clinical commissioning.
To manage the concern of maintaining a necessary network size, several sites developed smaller localities, clusters of practices within their network which resolve local issues, including commissioning. The localities’ leaders are typically part of the CCG board, responsible for leading the overall commissioning process. A GP commented, ‘[frontline engagement] won't work at three hundred thousand [patients] level […therefore] having those sub-groups, those cluster level groups is vitally important’ (site C). CCGs structure reflects the tension in achieving strong local commitment and efficiency through scale (see for a summary of the range in population size across study CCGs).
An important contextual feature that shaped the network size and its membership ties was the commissioning history, in particular, the legacy of PBC. Even though PBCs never held actual commissioning funds throughout their existence, they had established a distinctive ‘organisational archetype’22
which itself was a result of the sedimentation that took place during the organisational changes of the reform at that time. By and large, the specifications of the previous organisational archetype (in this case PBC groups) has an apparent effect on network formation and knowledge capability. In the reform process, change ‘represents not so much a shift from one archetype [PBC] to another [CCG], but a layering of one archetype on another’ (ref. 22, p.624), so that the new entity embodies the interlacing of previous structures and relationships with novel network features. As highlighted in our analysis of the vignettes around innovation between one former PBC and a non-PBC group, legacy ties between stakeholders influenced the innovation process.
CCG relationship with policy and administrative authorities
Another significant relationship influencing the new commissioning scheme is the relationship between GPs and health policy makers and administrators who oversee the implementation of the policy. Numerous GP leaders expressed frustration that a number of their colleagues are hesitant to engage because of the perceived weak engagement and lack of dialogue between policy makers (or their representatives) and CCG leaders. On the whole communication is seen as a one way process. During the course of the study we observed an increasing frustration among the CCG leaders. Several of them who were enthusiastic and motivated early on started to believe that their efforts were misplaced: ‘it was clear that there were many unfinished episodes and contradictions in the legislation, the Minister then turned to the professions in order to get their input and called those pathfinder organisations’. A GP from site A mentioned: ‘I was happy to contribute as a pathfinder under those terms but the pathfinders (forerunners of policy implementation) were used as evidence that the profession supported the Bill […] then I felt that I'd been tricked into being a pathfinder’.
As a result, numerous frontline GPs and CCG leaders commented they were becoming increasingly cynical and started questioning their engagement in CCG activities: ‘We're in between at the moment, waiting to know what the new world is going to look like, and not really being able to get on with things until that's clear’ (site A). In parallel with the uncertainty around the future of the reform contributing to network stability, CCG leaders felt that they have little guidance from the policy makers regarding their new activities and responsibilities: ‘the government is being less than explicit’. Yet at the same time CCG leaders did not feel able to shape the strategic direction nor develop new rules for the commissioning process, and this uncertainty was compounded by the simultaneous restructuring of PCTs.
Relational dynamics of early stage innovation in two CCG networks
In the vignettes below (see online supplementary boxes 2 and 3), we compare two interesting examples as to how relational dynamics in nascent CCG networks surrounding site A and site B enabled (and constrained) early stage innovation. We develop our insights concerning the relational dynamics drawing on the social network data ( and ) and the leadership challenges of working across the multiple stakeholders involved.
Site A (left) and site B (right) clinical commissioning group network diagrams.
In site A, where CCG leaders had access to comparative data from across the health system, the board leaders drew on existing strong relationships with the PCT to develop a solution in the form of joint working groups within the specialist areas and pathways of concern. The stimulus for the innovation process came from the available data highlighting the importance of network (in)coherence, coupled with the numerous ties with the PCT. As can be seen from the social network analysis comparatively illustrates the numerous ties among the CCG board and local health administration entities (PCT) in site A (18) which is higher than site B (13). The strong ties with the PCT was crucial in bringing together the other critical stakeholders (eg, acute providers) as the CCG board, had established ties with other stakeholders. In addition, the density of the ties across the board itself (0.737) indicates a high level of knowledge sharing and cohesion among the CCG leaders. This facilitated centrally coordinated action to develop the multiple pathway groups.
The social network diagram in illustrates the relatively uniform communication pattern across the board; it also brings to fore the very heavy reliance on a single knowledge broker (large blue node with high degree and betweenness centrality in site A). Overreliance on a small number of knowledge brokers adds risk to the network, for example in the case where the individual should exit the network. The network also becomes dependent on a few individuals who are able to commit a considerable amount of time for developing leadership processes.
Innovation emerged in site B from a frontline GP who recognised incoherence in one area of the network, given her knowledge of local primary-based care and specialist care. The board in site B is characterised by high levels of front-line GP engagement, illustrated both by the high numbers of direct ties to the board (6) and also the communication intensity between those ties, with relatively thicker blue lines in the social network diagram between board members and GP practices, as compared to site A. This enabled the innovation to be embedded and taken up by the GP community. However, as evidenced by the lower density of ties between CCG board members (0.622) there was an element of competition between the CCG leaders who represented the former PBC groups, indicated as the larger blue circles in the social network diagram (site B graph on the right of ). This influenced the integration and coordination of practices across the network as a whole, and hampered the scaling up of the innovative practice to other regions within the network.
In both cases, the development of novel care pathways arose from information regarding network incoherence, and a realisation that local care was out of alignment with care being provided in equivalent regions elsewhere. There was also a reliance on engaged frontline GPs and the use of strategically reconfigured knowledge flows to facilitate the development and delivery of a new service. Across the innovations new practices were knitted together from new relationships at multiple levels; structuring knowledge in new ways enabled novel insight as to how services could be integrated. Acting as relational catalysts rather than necessarily involving themselves in all relationship building, clinical leaders facilitated network coherence, stability and knowledge sharing in enabling innovations to emerge.