The capacity of health service commissioners to use evidence: a case study.

Article history

The research reported in this issue of the journal was funded by the HS&DR programme or one of its preceding programmes as project number 12/5002/01. The contractual start date was in September 2013. The final report began editorial review in December 2016 and was accepted for publication in August 2017. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HS&DR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

Sophie Staniszewska is a member of the National Institute for Health Research (NIHR) Evaluation, Trials and Studies Coordinating Centre Patient and Public Involvement Reference Group and is Associate to Professor Kate Seers (NIHR Health Services and Delivery Research Commissioning Board).

Disclaimer

This report contains transcripts of interviews conducted in the course of the research and contains language that may offend some readers.

Copyright statement

© Queen’s Printer and Controller of HMSO 2018. This work was produced by Currie et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2018 Queen’s Printer and Controller of HMSO

Conceptual problem

The research study was commissioned by the National Institute for Health Research (NIHR) Health Services and Delivery Research (HSDR) programme, under its knowledge mobilisation theme, following a number of other studies commissioned under the same theme in recent years. Such studies tend to empirically focus on the organisational level of knowledge mobilisation. Analysis and research-informed suggestions tend to lean towards the panacea of knowledge brokering (see, for example, a previous NIHR HSDR study by the same principal investigator, GC, 09/1002/051). Although knowledge brokering was proposed as a solution, Graeme Currie and colleagues criticised the concept and practice because it fails to account for the capacity of the organisation to mobilise knowledge. In this study, we extend the knowledge-brokering solution by accounting for the contextual intricacies that affect knowledge mobilisation. We enhance the evidence base for effectively mobilising knowledge through a system level of analysis regarding capacity to mobilise knowledge (e.g. of a complete commissioning network, which encompasses health-care providers and other organisations supplying commissioning intelligence). ‘Absorptive capacity’ (ACAP) represents such a concept, which can be applied at the health-care system level of analysis. 2 ACAP maps on to the term ‘critical review capacity’, and can be applied to consider the empirical case of health-care commissioning networks. In so doing, our study addresses a need for the development of a model of ACAP that is more tailored to health care, a need that was encompassed within the original NIHR HSDR call for commissioned studies. 3

We highlight studies of ACAP in health care that are already published. 4–6 We seek to develop a model of knowledge mobilisation focused on ACAP of health-care commissioning networks, but one that is more widely transferable across the health-care system. Within our study, we draw on a more generic literature to set out some of the expected characteristics of health care that affect dimensions of ACAP and its antecedent combinative capabilities. 7–13 We then consider distinctive features of the health-care context that may affect ACAP.

Applying the concept of ACAP, we argue that knowledge mobilisation occurs within overlapping processes, which can be conceptualised as acquisition, assimilation, transformation and exploitation of knowledge. These processes are influenced by combinative capabilities: systems capabilities, socialisation capabilities and co-ordination capabilities. Systems capabilities, such as targets and incentives and standardised data sets, can limit the type of knowledge acquired and used to guide service interventions. Socialisation capabilities, represented by professional and organisational cultures and associated professional–managerial relations, can limit knowledge mobilisation across occupational groups. Co-ordination capabilities can overcome the barriers of systems and socialisation capabilities, encouraging more flexible approaches to the four stages of knowledge mobilisation. In particular, this study highlights the importance of general practitioner (GP) involvement, patient and public involvement (PPI), business intelligence (BI) and social integration mechanisms, which further enhance knowledge mobilisation in health-care settings.

Practical problem

Our study provides help with the challenge the NHS faces in the wake of recent structural reform, but also supports the long-standing concern of policy-makers and research institutions, including NIHR, about the ability of the NHS to translate different types of evidence into transformative action for service development.

With respect to the long-standing translation challenge for the NHS, the development and validation of an ACAP model contextualised for the health-care system offers considerable promise to address knowledge mobilisation at the levels of individual, group and organisational behaviour, as well as systems-level behaviour. Particularly in financially parsimonious times, Quality, Innovation, Productivity and Prevention (QIPP) gains are likely to be realised through better mobilisation of existing knowledge from both formal and local informal sources, rather than necessarily generating new knowledge. A model of ACAP contextualised for the case of the health-care system will guide action regarding how this might be realised. The case of intervention to reduce potentially avoidable elderly admissions into hospitals represents an exemplary tracer study revealing lessons for other complex areas.

Recent government policy in England emphasises that services in the NHS can be improved through the increased involvement of clinicians and service users at local levels. These policy developments have resulted in the creation of Clinical Commissioning Groups (CCGs), devolving financial decision-making and commissioning of services to clinicians based in general practice, and thereby giving them greater freedom to respond to the needs of local client groups. 14,15 Our study is particularly timely in the face of NHS commissioning reforms, but we note that knowledge mobilisation in commissioning is a long-standing concern evident in previous primary care trust (PCT) arrangements, and is likely to extend to any other commissioning structures and processes in the future. In short, our study is intended to produce and disseminate analytical lessons regarding the enhancement of critical review capacity beyond current arrangements, and with a focus on the wider commissioning network, not merely on CCGs as organisational entities. Our analytical lessons also extend beyond the tracer study of elderly care. We suggest that the challenge of commissioning high-quality elderly care in a productive but preventative way is exemplary of a more general challenge of commissioning complex care, which often encompasses more than a single disease pathway and crosses boundaries within and outside the NHS.

With respect to current structural reform, should CCG-led commissioning networks fail to develop critical review capacity (which our study conceives as ACAP) for commissioning, structural reform is unlikely to succeed. Commissioning networks developing critical review capacity depends on enhancing antecedents for ACAP, that is, understanding how combinative capabilities of the commissioning network affects ACAP. We emphasise that our study is concerned not just with CCGs, which represent the centrepiece organisations in current reforms, but with the commissioning network as a whole, inclusive of a wide range of organisations. So, if CCGs were to be supplanted in a future round of structural reform, our focus on the commissioning network as a whole means that our study still holds practical relevance. Our stance regarding the wider commissioning network, rather than focusing on CCGs alone, is supported by Swan et al. ,16 who, in their study of previous commissioning arrangements, highlight the multilevel interdependencies around commissioning that mean that any commissioning organisations, such as CCGs, need to build effective relationships with, for example, the NHS Commissioning Board, including their local area teams; GPs and their practices; public health; local authorities; commissioning support units (CSUs); acute hospitals and other providers, such as community health, mental health and care homes providers; and other bodies that may provide commissioning intelligence, such as out-of-hours services, emergency services, the Care Quality Commission and Healthwatch England.

Research examining critical review capacity related to the new commissioning arrangements is scarce as a result of how new the organisational networks are. Limited research that does examine new commissioning networks highlights a need to increase critical review capacity for evidence-based decision-making through not just acquiring knowledge, but also using that knowledge. 17

Specifically, CCGs struggle to commission effective, targeted interventions for reducing potentially avoidable elderly care hospital admissions. Our study will indirectly reduce some of the uncertainty that hospitals face and that manifests itself in crises such as a lack of hospital beds in winter, as commissioning networks improve their ability to mobilise knowledge to commission interventions to prevent potentially avoidable elderly care admissions to hospital.

Regarding patient benefit, there is strong evidence that patient/carer experience of care and clinical outcome improves through being kept out of hospital. For the NHS, potentially avoidable admissions prove to be difficult and costly to manage. Frail elderly admissions, many of which can safely be avoided, are both clinically and economically deleterious. 18,19 Noting considerable variation in admissions, it is estimated that bringing provision up to best practice in England could result in savings of £132M,20 and a reduction of 7000 beds. 17 Proactive integrated care can reduce the proportion of frail elderly patients who go to hospitals. 21,22 Good care is diligent care that recognises and responds to signs of deterioration in early stages. Social support in the community is also important for people with both physical and mental impairment, with approximately half of all elderly patient admissions fulfilling the criteria for dementia. 23 Evidence might be used to commission community care that is sensitive to clinical pathways, recognises that most people have comorbidity and reduces potentially avoidable admissions of older patients. 19,24

Taking account of the above, we empirically focus on the service domain of elderly care, specifically a tracer study of capacity to mobilise knowledge to reduce otherwise avoidable admissions into hospital. This, as highlighted above, is particularly relevant to the case of ‘frail’ older people, a population group that our research sought to focus on in interviews with key stakeholders in CCG-led commissioning networks. However, as evident in empirical material, there is some slippage in our characterisation of the group we seek to reduce avoidable admissions of. This reflects the varied capacity of our 13 comparative cases of commissioning networks to acquire and use knowledge for intervention targeted at frail elderly patients rather than the larger, more general, group of older patients (over-65s who are not ‘frail’). We apologise in advance for what might appear to be imprecision, but argue that it derives from empirical data that we wanted to use effectively.

In considering a commissioning network’s acquisition and use of evidence, it is not our intention to set priorities for CCGs or to say how different sources of evidence (e.g. academic research, voluntary sector, PPI, GPs’ experiential knowledge and BI) might be combined. We suggest that different forms of evidence should be deemed to be of different levels of importance depending on the decision to be made, but nevertheless do not go about assessing this. What we would say is that ‘effective’ commissioning demands that CCGs acquire and use pluralist types of evidence in making decisions about service interventions. We stress that, although it is one thing to acquire evidence, such as that about the patient experience, it is another thing to use such evidence. Effective commissioning demands both acquisition and use.

Research question

Our conceptual contribution is the development of a model of ACAP encompassing its four dimensions (acquisition, assimilation, transformation and exploitation) and its three antecedent combinative capabilities (co-ordination, systems and socialisation), which are applicable to a wide range of health-care organisations. Our overarching empirical research question (RQ) is: how can we enhance ACAP of CCG-led commissioning networks in health care to inform decisions to reduce potentially avoidable elderly care acute hospital admissions?

Research methods

We adopt mixed methods (mainly semistructured interviews, but also PPI focus groups, observation, secondary data and, in the future following validation, the ACAP psychometric tool) to examine comparative cases of commissioning networks regarding external and internal contingencies around their critical review capacity and to conceptualise critical review capacity as ACAP. See Chapter 3 for more detail.

Study governance

A strategic advisory group governed the study (meeting four times throughout the course of the study). Alongside this strategic advisory group, a distinctive PPI group was formed, which met regularly (also meeting four times throughout the course of the study). Our thanks go to Tony Sargeant (Chairperson of the strategic advisory group and PPI group) and his colleagues, who, beyond providing the necessary governance for our study, provided invaluable advice that shaped our empirical endeavour and, in particular, helped develop and provide face validation for the ACAP tool. The PPI group’s reflections on our study and its findings are presented in Chapter 10.

Structure of the report

We commence with a literature review, which outlines and critiques the ACAP concept before discussing its recent application to health care (see Chapter 2). Following this, we outline our research design (see Chapter 3). We then set out our empirical findings across five chapters: constraints of socialisation and systems capabilities (see Chapter 4), GP involvement (see Chapter 5), PPI (see Chapter 6), BI (see Chapter 7) and social integration (see Chapter 8). This is followed by a chapter that discusses the development of the ACAP tool, exhibits its dimensions and suggests how it might best be used (see Chapter 9). We then present our PPI group’s reflections on our study and its findings (see Chapter 10). Finally, we have a concluding discussion, within which we highlight our contribution to cumulative knowledge, practical implications, transferability of findings and further research (see Chapter 11).

Introduction

Recent reforms in NHS England have seen the introduction of CCGs, which lead commissioning networks, replacing the previous commissioning structures of PCTs (see Timmins25 for an in-depth overview of the policy progression from PCTs to CCGs). It is proposed that CCGs will increase positive clinical outcomes, financial management and service effectiveness through the creation, assimilation and exploitation of existing clinical guidelines and new innovations. 15,26,27 For these processes to take place, CCGs will need to critically review the external information they receive before assimilating, transforming and exploiting this knowledge through their commissioning decisions.

Given that CCGs are new organisations, literature on their critical review capacity for intelligent commissioning is sparse. Nevertheless, early scoping studies highlight the weak critical review capacity among CCGs,17 which was also apparent in their predecessors, PCTs. 16,28 One study produced a live simulation with a number of new CCGs to explore how they would adapt to some of the challenges they may face in the next few years. 17 The results indicated some areas of concern as GPs leading the groups struggled to work in a proactive, innovative way, acknowledging that they were overwhelmed by transactional demands that reduced their critical review capacity. These concerns have been echoed more recently, as researchers question whether or not commissioning reforms have actually led to increased performance and quality in health-care delivery. 29

Studies highlight that the quality of services delivered by health-care organisations is improved when the organisational capacity for learning and integration of diverse sources of knowledge is developed. 30–32 Recently, academic studies have highlighted how the concept of ACAP,2 derived from private sector settings, can be applied to health-care organisations to address how they might enhance service interventions by improving their critical review capacity. 3–5,33,34 In this chapter, we outline the concept of ACAP, which describes an organisation’s ability to acquire, assimilate, transform and exploit knowledge from the environment, improving competitive advantage2,13 in the context of CCGs. We highlight that the realisation of an organisation’s critical review capacity comes from reducing the variance between its potential absorptive capacity (PACAP) and its realised absorptive capacity (RACAP). Following this, we explore the antecedents to ACAP, known as combinative capabilities, highlighting how they can enhance or stymie the critical review capacity of organisations. 9–12 Finally, we highlight the importance of four co-ordination capabilities for enhancing the critical review capacity of CCGs.

Absorptive capacity

The concept of ACAP was first coined by Cohen and Levinthal2 to describe an organisation’s ‘ability to identify, assimilate, and exploit knowledge from the environment’. Lane et al. 9 hail ACAP as ‘one of the most important constructs to emerge in organisational research in recent decades’. Since its inception, ACAP has been seen as a core element of increasing critical review capacity, but the dynamics of enhancing ACAP are theoretically underdeveloped. 6,9

In this study, we adopt Zahra and George’s13 conceptual framework of ACAP, which characterises four activities of pulling in external knowledge that are crucial for the development of ACAP: (1) identifying and accessing relevant knowledge through acquisition processes; (2) analysing and interpreting this knowledge through assimilation; (3) integrating existing knowledge with the newly assimilated knowledge through transformation; and (4) refining and developing existing organisational routines and behaviours through exploitation of the transformed knowledge.

We recognise that health care represents a distinctive environment compared with private sector research and development (R&D) contexts in which much of the empirical work around ACAP has taken place. This renders some of the dimensions of the ACAP literature less relevant to health care, but, at the same time, brings others to the forefront. 5 In particular, in application of ACAP theory to literature, it is important to recognise two interacting and complementary elements: PACAP, the ability to acquire and assimilate knowledge; and RACAP, the ability to put newly acquired knowledge into action within the organisation through transformation and exploitation (as outlined in Figure 1). It is the variance between PACAP and RACAP that accounts for performance differences among organisations, with a decrease in variance crucial to enhancing ACAP. 13 In the following section, we discuss PACAP and RACAP in the context of CCGs.

FIGURE 1.

Model of ACAP (adapted from Zahra and George, 200213).

Reducing the variance between potential absorptive capacity and realised absorptive capacity: the influence of combinative capabilities

It would appear from the literature review that any interventions aimed at increasing the effectiveness of CCGs need to be focused on increasing RACAP. The ability to transform and exploit knowledge throughout the organisational network, and to implement innovations and recommendations, is a fundamental need for CCGs. Consequently, when considering critical review capacity in health care, there is a need for research into how organisational antecedents affect ACAP, which takes into account organisational context, the role of individuals and groups, and associated power and politics. 8,9,12,13 This requires more focus on social mechanisms, as the CCG will need to develop the ability to share and transform information and knowledge to make critical decisions. The difficulties associated with knowledge sharing and transfer cannot be resolved through redesign of governance structure because of the socially embedded nature of tacit knowledge, which is a key element of critical decision-making in social contexts. 50

Although existing research into ACAP has engendered deeper understanding of the organisational conditions that have an impact on organisational knowledge processes, literature in this theoretical domain remains focused on knowledge acquisition, rather than its use,6 and there is a lack of detailed understanding of the antecedent capabilities for enhancing ACAP, and in particular of reducing the variance between PACAP and RACAP. 8,11 What is clear is that there are antecedents to ACAP, specifically known as combinative capabilities. 10,12 As noted previously, health-care organisations represent a distinctive context compared with private sector R&D contexts, in which much of the empirical work around ACAP has taken place, and, as such, combinative capabilities will influence the four ACAP mechanisms in health-care organisations in a particular way. 5 Van den Bosch et al. 11 distinguish three types of combinative capabilities that influence ACAP: socialisation capabilities, system capabilities and co-ordination capabilities. In the following section we discuss these capabilities in the context of CCGs.

Co-ordination capabilities in commissioning

Identifying and enhancing co-ordination capabilities is particularly important in complex organisational settings, such as health care, ‘where . . . the ability to engage in flexible, knowledge exchange and generation across hierarchy and function – are essential to organisational performance’. 33 However, as noted previously, the influence of intraorganisational dynamics on co-ordination capabilities has received little attention in the existing literature, within both health care and the wider body of organisational research. 12 Although dimensions of co-ordination capability have been identified within health-care settings,3,4,32 there is a lack of understanding of how such co-ordination capabilities work to enhance ACAP, and how development of co-ordination capabilities might be supported. Drawing on the ACAP literature outlined above, it appears that there are key elements of co-ordination capabilities that could be drawn on to improve the critical review capacity of CCGs: information support systems, social integration mechanisms, clinical participation in decision-making, appropriate leadership and, more generally, social relations with diverse group inside and outside the organisation. 33 In the specific context of CCGs, we conceptualise these co-ordination capabilities as four mechanisms: GP involvement, PPI, BI functions and social integration mechanisms.

Conclusions

Absorptive capacity is a theoretical concept that can be used to explore critical review capacity in organisations. 2 Of particular importance is the consideration of variance between PACAP (acquisition and assimilation) and RACAP (transformation and exploitation), an area that has been highlighted for future research. 8 Although the majority of research concerning ACAP is limited to the private sector, in this chapter we reported the potential benefits when applied to the health-care context of CCGs, outlining previous research that suggests that health-care organisations might apply ACAP concepts to enhance service interventions by improving their critical review capacity. 3–5,33

In particular, we highlighted the important, yet underdeveloped, effect of combinative capabilities on the reduction in variance between PACAP and RACAP. Indeed, Van den Bosch et al. 11 call for future research to assess not just ACAP stages but how combinative capabilities affect the acquisition, absorption and use of new knowledge for improved organisational performance. The literature suggests that systems and socialisation capabilities act to reduce ACAP in a more pronounced way in health-care organisations. 8 This highlights the need for health-care organisations to develop co-ordination capabilities to offset the effects of systems and socialisation capabilities so that their ACAP is enhanced.

The four co-ordination capabilities that are identified as important in the context of CCGs are GP involvement, PPI, BI functions and social integration mechanisms. Each of these co-ordination capabilities has the potential to mediate some of the barriers of systems and socialisation capabilities, which characterise complex organisational contexts. However, although policy identifies the development of these four elements as a priority for CCGs, the extent to which each is able to reduce the variation between CCG PACAP and RACAP, enhancing critical review capacity, is unclear.

Introduction

To explore how CCGs can enhance their critical review capacity for intelligent commissioning, we followed a tracer study:106 that of commissioning interventions to reduce avoidable admissions of older persons into hospitals. For the NHS, potentially avoidable admissions prove difficult and costly to manage. Frail elderly admissions, many of which can safely be avoided, are both clinically and economically deleterious. 17–20 Avoiding admissions is a matter of improving care in the community, which is particularly important for people with both physical and mental impairment, with approximately half of all elderly patient admissions fulfilling the criteria for dementia. 23

Elderly care admission avoidance was therefore chosen as a tracer study owing to its central importance, particularly in the face of the austere financial climate facing health care. As such, the study contributes to system transformation that takes costs out of hospitals, and links to local transformational efforts around QIPP, a significant part of which focuses on system improvement and cost reduction in elderly care. Our study aims to reduce some of the uncertainty that hospitals face and that manifests itself in crises, such as a lack of hospital beds in winter, as CCGs improve their ability to anticipate fluctuations in demand and commission interventions to prevent potentially avoidable elderly care admissions to hospital. Improvement in organisational and clinical outcomes is enhanced by our consideration of joint commissioning across health and social care, noted as particularly challenging,96,105 with Swan et al. ’s16 2012 study of PCTs further highlighting complex, multilevel interdependencies in commissioning decisions across organisations. Practically, we suggest that the relationship between CCGs and acute hospitals may be prone to tension in this area as productivity gains ensue through transfer of care outside hospitals, so, in this regard, the tracer is interesting.

Further to the theoretical reasoning for selecting elderly care admission avoidance as the tracer study, the issue was viewed as a particularly pressing issue by the CCG leads whom we interviewed in an exploratory study. However, the results from this study are not limited to care of older people, as application of the ACAP concept allows theoretical generalisation to other commissioning domains. 107 In other words, although we focus specifically on commissioning processes related to elderly care, the underpinning processes and findings are generalisable to a broad range of commissioning issues.

Our overarching RQ is as follows: how can we enhance ACAP of CCG-led commissioning networks in health care to inform decisions to reduce potentially avoidable elderly care acute hospital admissions? This breaks down into three linked RQs:

1. How do CCG external relationships and system antecedents to new commissioning arrangements affect CCG critical review capacity?

2. What features of CCG organisational context (i.e. internal capacity) affect CCG critical review capacity?

3. How do we reduce the variation between PACAP and RACAP?

Case studies

We follow a comparative case study approach,108 focusing on 13 CCGs in NHS England, allowing generation of more robust analysis, alert to contextual contingencies, when considering the antecedents of critical review capacity. Case study research involves a multilevel approach to the understanding of in-depth, dynamic settings, and can expand knowledge of individual, group, organisational and social phenomenon by allowing researchers to examine real-life events in a rich, holistic context. 107 Case study research does not attempt to select statistically representative samples, instead focusing on events or programmes that provide a unique conceptual insight into a theoretical framework at different levels of analysis. 107,109,110 Interviewing participants from different hierarchical levels and organisational positions will provide different perspectives on phenomenon, an essential element of multiple case research. 111,112

Case study research explains processes such as how we enhance CCG critical review capacity. Comparative case analysis allows generation of more robust analysis, particularly when concerned with drawing out contingencies, such as those external and internal contingencies framing CCG critical review capacity. 111 Comparative case analysis also accommodates heterogeneity of CCGs, and our study sites are representative of heterogeneity nationally.

The eventual sample of 13 CCG-led commissioning networks is representative of their characteristics (Table 1); however, we faced a considerable challenge in accessing empirical sites, that is, we approached many more potential empirical cases. Commonly, we experienced some interest and an invitation to discuss our study with a key stakeholder, often the chief operating officer, but were then passed down the CCG organisation to a GP commissioning lead or commissioning manager. Beyond the chief operating officer, other stakeholders sought to translate our research intent towards something else of more operational priority (at which point we politely disengaged). Our overall impression was that CCGs remained rather passive about developing their critical review capacity, even though, in our eyes, this was necessary to achieve their strategic remit set out by policy-makers.

However, once we negotiated access to our 13 empirical cases, we asked each CCG to suggest a commissioning issue related to elderly care admission avoidance that was currently important in their organisation. The focus of each case fell broadly into the four categories of co-ordination capabilities highlighted in the literature review (see Chapter 2): GP involvement, PPI, BI functions and social integration. The demographic spread of each CCG, and their specific commissioning focus discussed during interviews, is outlined in Table 1.

Interviews

Comparative case studies were underpinned by qualitative fieldwork, with data gathering mainly via 159 semistructured interviews,113 most of which were conducted by the fieldworker (research fellow employed on the study), with the principal investigator for the study conducting a smaller number to ‘keep in touch’ with the primary data. Interview schedules were informed by themes generated from the literature review, which was thematic and followed narrative models. 114 In-depth interviews allow researchers to understand how the respondent views the world without researcher preconceptions being imposed on them, and to generate rich data with minimal researcher reactivity. 115,116 Individual interviews allow an insight into how large-scale transformations are experienced and are affected by the interactions of individuals, and they focus on how these interactions are embedded in the social and cultural context. 117 Semistructured interviews offer a much richer account of how the interviewees’ experiences, knowledge, ideas and impressions are considered and documented. Interviewees are less constrained by the researchers’ pre-understanding and there is space for the negotiation of meanings, so some level of mutual understanding is reached. 118

Our sample of interviewees from the 13 CCG-led commissioning networks represented stakeholders who were seen to be central to the commissioning process, and in many cases carried some ‘managerial’ responsibility for commissioning. We also engaged with ‘lay members’ of the CCG, individuals without clinical backgrounds who participated in PPI groups, or who acted in a lay member capacity on the CCG governing body. With assistance from the relevant CCG chief operating officer in exploratory interviews designed to engage CCGs in our study, we identified some respondents a priori, and then followed a snowball sampling pattern119 until the themes emerging from interviews were theoretically saturated. To ensure that the extended commissioning network was considered, interviews were not limited to those working within CCGs but also encompassed those working in hospitals, in addition to public health and local government professionals concerned with provision and commissioning of older people’s care. Further to this, to reflect the ongoing politicised nature and top-down control that remains within health-care contexts, we interviewed six NHS England staff, including three of the NHS Commissioning Board local area team leads linked to our focal CCG empirical cases. After we obtained ethics approval from local NHS research ethics committees and R&D departments, a total of 159 participants were interviewed (Table 2).

Within each CCG network, we undertook semistructured interviews and asked respondents to describe the commissioning process, focusing on the four processes encompassed within knowledge mobilisation, embedded in an organisation’s ACAP (acquisition, assimilation, transformation and exploitation), and their antecedents or combinative capabilities. We did not directly invoke technical terms such as ACAP and capabilities, but asked more general questions, such as the following: ‘How do you acquire data and information about elderly care hospital admissions?’; ‘How do you use such data and information?’; ‘What are the barriers to using data and information?’; and ‘How are these barriers mediated?’. Participants were also encouraged to speak freely about their experiences of commissioning processes, and interview themes were guided by issues arising during the conversation.

As we gathered data, we sought to collect rival explanations for the phenomena reported, in a spirit of scepticism about reporting of events and actions. We sought to collect discrepant evidence to develop a plausible rival description of events and activities. In short, we sought to challenge our explanation of knowledge mobilisation by CCG-led commissioning networks as one focused on ACAP. There were specific explanations of poor use of knowledge mobilisation by CCG networks. More generally, top-down policy pressures, power differentials with large acute providers and bureaucratic tendencies of CCG employees converged, meaning that knowledge mobilisation was poor. Theoretically, the concept of ACAP provides an explanation for poor knowledge mobilisation.

Patient and public involvement during the study

Although the overarching study was concerned with enhancing decision-making processes for commissioning organisations, specifically related to interventions to reduce avoidable admissions of older people into hospital, we were also interested in PPI in CCG networks. Although only three of our networks explicitly asked us to explore PPI, it was a central component of our interview schedule at each site. Over the course of the study, distinct variations in managerial interventions shaping PPI involvement were noted, leading to different outcomes for PPI influence on service development at each site. Although all commissioning networks had formalised structures for public engagement, the managerial influences on PPI varied immensely (this is discussed in detail in Chapter 6).

In addition to questions relating to ACAP, as outlined above, we also asked interviewees to describe how information or opinions were acquired from the public and through what structures this took place; how feedback was used with other forms of data to determine the needs of the local community; how information from PPI was used to design services; and what the influences on PPI were. In addition, when possible, the research team intended to attended public engagement forums or meetings at each site. This was limited, in most cases, by managerial reluctance for us to observe meetings and, for this reason, was restricted to six observations of PPI meetings occurring over three sites (those specifically interested in their public engagement). Extensive hand-written field notes were taken during observations and used to supplement analysis of themes related to PPI.

In addition to the focus on PPI during research, the research group itself had a strong PPI focus. The study advisory board, which meets yearly, is chaired by Tony Sargeant, patient involvement champion. In addition to the research staff on the board, Andrew Entwistle, lay specialist in patient involvement, and Graham Martin, academic specialist in patient involvement, were key members of the study advisory board. This group was heavily involved in the design of the study, the ongoing planning and response to themes emerging, the analysis of the findings and the dissemination plans for the findings. Further to this, Tony and Andrew also sat on the study PPI group. This group met at least twice a year to receive regular study updates from the research team and to triangulate study analysis. They offered insight into the direction of research and highlighted issues that the research team may not have isolated as influences on critical review capacity in CCGs. Finally, the PPI group were also involved in the dissemination of research. Crucially, the ACAP psychometric tool was coproduced between the researchers and PPI group in an iterative manner (see Chapter 10 for further details). Their meetings were audio-recorded by Sophie Staniszweska (co-investigator), who continues to work with the group to produce some user-led publications or study outputs.

Analysis

Semistructured interviews lasted between 45 minutes and 1 hour and were audio-recorded and transcribed. Initial coding was carried out by one member of the research team (research fellow) following template analysis. Coding was then discussed with the principal investigator, with the research fellow and principal investigator jointly engaging in further rounds of coding as we moved towards more abstracted theoretical themes. Coded material was organised in a hierarchical manner to provide general, parallel and subcategory codes to allow analysis at different levels of specificity. This allowed a flexible and adaptive approach to research analysis, with some a priori codes drawn from literature review, but with the detail of these and how they inter-related induced from analysis of the data. 120,121 Observational field notes and documentary evidence from the CCG networks helped generate early contextual understanding of empirical sites,122 but the data presented draws solely on interview transcripts. Coding was facilitated through the use of the software program NVivo 11 (QSR International, Warrington, UK), which produced a more rigorous and transparent approach to data analysis. 123

We checked our emerging analysis with the study PPI group, research team and study advisory board, and through iterative feedback with each CCG site, to inductively identify how the four co-ordination capabilities (GP involvement, PPI, social integration and BI functions) were evident in interview transcripts. This was then refined by coding for examples of the three combinative capabilities: barriers of systems capabilities, barriers of socialisation capabilities and co-ordination capabilities as a mediator. An example of the coding structure is presented in Table 3. Following this, we present the findings of each of these co-ordination capabilities in detail.

Limitations

A common criticism of case studies, even when there as many as we sampled, is their uniqueness and their ‘generalisability’. Although we attempted to elicit a representative sample of CCG-led commissioning networks for study, they proved difficult to access and many of the initial agreements for access fell through as key personnel acting as gatekeepers changed, as we were passed on to other staff who were to operationalise the study and as staff in CCGs came under increasing work pressures. Therefore, some readers might judge our sample of case studies as less representative along other dimensions of commissioning networks; for example, given the empirical tracer of elderly care, we might have sampled for population demographics. Nevertheless, we have attempted to generate transferable lessons for other commissioning networks through theoretical analysis, drawing on dimensions of ACAP to highlight factors influencing knowledge mobilisation.

Similarly, as revealed in our table of interviewees, although managerial staff engaged with our research, accessing GPs proved more challenging. Of all commissioning stakeholders, they are subject to the greatest work pressures, particularly when they undertake hybrid roles combining commissioning and clinical responsibilities. We thank those GPs in particular who gave up their valuable time to participate in our study. However, given the variable influence of different groups of GPs revealed within our empirical analysis, we encourage others to engage in more specific study of GP involvement in commissioning to ascertain if this has spread beyond a narrow cadre of GPs who engaged similarly with previous arrangements under PCTs.

At the inception of our study, CCGs were fledgling organisations. As highlighted in empirical analysis, the development of social and organisational relationships across the commissioning network takes time. Our study only partially revealed this given the limited time period over which we undertook data collection. Others may follow our study, perhaps revealing the development of critical review capacity of CCGs as their relationships developed across the commissioning network.

Systems capabilities

Systems capabilities differed slightly between CCGs in terms of the IT systems used to gather information, but all had access to similar data sets, guided by the same centrally determined governance structures:

We have our own information . . . what patients have gone to hospital, what’s happened to them. Also we have a lot of information for secondary care. We can look at trends, emergency admission trends, elective admission trends. For long-wait domains we actually have an absolutely granular understanding of waiting lists to make sure we’re going to hit key targets.

CCG B, interview 1

The thorough nature of secondary care data sets was frequently acknowledged as enabling them to be particularly strong sources from which to acquire information. However, the ease with which CCGs could assimilate and transform these data into commissioning decisions was not as clear. Many respondents voiced concerns about their ability to use the information to fully guide their decision-making:

There’s an awful lot of data out there, but it never really comes back as intelligence and I see quite a lot of leaps of faith in terms of data to conclusions.

CCG C, interview 4

This problem was also articulated by a manager from another CCG, who noted ‘we are awash with data, but probably not awash with information’ (CCG B, interview 2). One reason for this barrier to assimilating the data was related to the centrally predetermined nature of the information, which was not always meaningful in terms of its representation of the CCGs’ local population needs:

It’s not geared up around us as a CCG. It’s not really meaningful for us in terms of making decisions.

CCG C, interview 3

This was further illuminated by respondents who acknowledged the benefits of using local information to commission in response to patient needs, but also reported barriers to doing so. For example, one GP commissioner was clear on the need for CCGs to commission according to local need:

You can’t use national statistics to drive what we’re doing locally because our patients have different needs and we need to build data up from the ground asking ‘Well, what do we need for our patients? How should that look? How will it serve them best from a clinical perspective?’.

CCG A, interview 7

However, that same GP commissioner subsequently described the problems in accessing that information:

So one of the barriers we’ve met right now is we’re trying to get the data for local patients and it seems to be a nightmare trying to get the data. All sorts of people hold this data and say ‘Oh well, for governance issues you can’t have this and you can have this. You can have the regional data, but not the local data’.

CCG A, interview 8

In addition to the concerns about the local responsiveness of data, there was a sense of distinct lack of information from community or tertiary services. This was a particular problem for commissioners attempting to reduce potentially avoidable elderly care admissions into hospital by ensuring comprehensive community care. Although secondary care providers could generate information about admissions, interventions and discharges, community services could only provide vague metrics that were not always seen as helpful for the commissioning process:

But for us as commissioners even if we have the data, we need to know is that a good thing or a bad thing? Well I’ve no idea. It doesn’t mean anything. We have information about different services provided . . . but the information is fundamentally a garbage metric whichever way you look at it.

CCG B, interview 1

Subsequently, the systems capabilities of CCGs were undermined by their reliance on data sets that did not provide comprehensive information about their service provision. This was also compounded by continuing central control, which could undermine the ability to commission according to local needs, as CCGs also needed to conform to centrally set priorities:

The National Commissioning Board have started saying ‘These are your targets. You’ve got to do this. This is how you will do it’, and they’re starting to be very prescriptive.

CCG C, interview 7

However, despite the problems undermining the systems capabilities of CCGs, it was evident that there were also mechanisms in place to mediate these problems and enhance their capabilities. Most notable was the integration of clinical knowledge, through GP involvement, into the system:

They’re watching us really closely, but it does also still feel like we are in control . . . we listen to them, of course we do . . . but we’ve still got to use our patient data and everything that we know about them to drive intervention.

CCG A, interview 5

This was also acknowledged by a member of the management structure, who reported that the gaps in information with regard to service provision were less important when commissioning decisions were guided by the tacit knowledge of clinicians. This increased their systems capabilities and, subsequently. their commissioning capacity:

If I’m being honest, there’s probably some retro fitting of that information to the things that the GPs want to do. So kind of regardless of all of that analysis, from the experience of the GPs in their consulting room the key issue is with the elderly and we need to be more integrated.

CCG C, interview 1

Therefore, the involvement of clinicians within the data acquisition process could mediate the problems encountered as a result of their limited systems capabilities. However, unleashing the potential of that information also relied on CCG socialisation capabilities.

Socialisation capabilities

Socialisation capabilities mediate the extent to which CCGs are receptive to new forms of knowledge or information. The importance of developing a shared identity and common vision was repeatedly noted by respondents:

You’ve got to get people to own the problem and feel like it’s theirs . . . It’s one of those things where you need to take the time to build the relationships because you’ll get the payback if you do.

CCG C, interview 6

However, this is problematic as CCGs are highly professionalised contexts with two distinct cadres: managerial staff and professional staff (GPs). Although they are often referred to as new organisational forms, they are based on pre-existing organisations that have historical footprints. Indeed, the way many CCGs were set up created the potential for repetition of institutionalised behaviour, as managers were recruited straight from pre-existing structures:

We had a chief exec[utive] at the time and his view was that he didn’t like uncertainty, so when the policy about CCG creation came out he wanted to very quickly move towards new CCG structures. So basically he approached the old practitioner-based commissioning clusters which were there before and they were largely on the old footprint of the previous PCTs . . . and said ‘Right OK, you need to find yourself a chief officer and then that person needs to get some structures in and then we’ll transfer people from the PCT into these structures. Get on and do it’.

CCG B, interview 2

The direct transfer of management staff between old and new structures was seen to exacerbate institutionalised behaviours, with some GPs reporting that, on occasion, managerial behaviours had been reminiscent of previous organisational structures:

Initially I think some of the key managers didn’t really get what we were trying to do . . . For example, one of the CCG managers came up with an idea and just e-mailed all the GPs saying ‘This is what we’re going to do’, and we said ‘Hold on a minute! Where’s the discussion about this? You’ve not even asked what we think.’. So it then actually went back a good three or four steps to engage with GPs, engage with nurses, engage with the stakeholder reference group, then bring it to the governing body to have a further discussion before we went ‘Yes, this is now going to happen.’. Because in the old days that’s all we used to get – the PCT e-mailed round going ‘You’re going to do this’.

CCG A, interview 6

The institutionalised behaviours linked to the previous organisational forms were seen to undermine CCG capabilities, as the GPs sometimes reported a sense that managerial staff had different priorities to them, generating an organisational divide:

I think a lot of clinicians are a bit dismissive of managers . . . the perception is that nurses and doctors in hospitals are the heroes and managers are sort of boo, hiss, pantomime villains and that we shouldn’t have any managers . . . the sense of putting patients first is very important. We don’t tend to lose ourselves in sort of targets and things as I think some managers do.

CCG C, interview 7

This exacerbated tensions between managers and GPs, and was reported by managerial staff as well as by clinicians:

There’s tensions that are starting to come out between the managers and the clinicians . . . I think they do think as managers we sit there twiddling our thumbs quite a lot of the time and there is a tendency to sometimes feel like they’re trying to run it like a practice, and I don’t think they really know what clinical leadership means . . . doctors’ egos still get in the way to be perfectly honest . . . There’s no sense of ‘we’re a team’.

CCG C, interview 2

The interaction between managers and GPs, and the perceived obstructiveness of both groups, had a negative impact on the cohesiveness of the CCG. Generating a sense of being a ‘team’ was acknowledged as something that was going to take some time, particularly in CCG C, which had encountered some divisions between its managerial and GP staff (as outlined above):

It’s a lot of hard work to build a Board, to build the cohesion, the understanding between people, the mutual support and the mutual challenge within that so you build a team and building a team is not something you do overnight and we’re only just embarking on that process.

CCG C, interview 9

However, CCG B seemed to have overcome the majority of tensions between the different groups by investing time in CCG team development and setting a joint direction:

We’ve got a really good governing body. We’ve really gelled together as a team. That’s really important. We get on well, but we also get the job done . . . There’s clearly a culture here. There’s very much a ‘yes we can do, we do want to change, we want to be different’ culture from the top down and I think that’s clearly what makes a difference if you’ve got an inclusive management style.

CCG B, interview 8

Despite this, influences on socialisation capabilities were not limited to the immediate CCG context, but extended to the wider network. Our informants perceived that secondary care providers may not be working towards the same vision. This may have been the case with other network partners as well, but the large size and influence of secondary care meant that discussions about these relationships often came to the fore during interviews:

The acute services have this huge gravity which pulls in commissioning thinking . . . The Trust has a contract of more than £600M so that’s six times [our] entire commissioning budget, so it’s a vast organisation and a very powerful, huge power in the marketplace.

CCG A, interview 3

This power differential was problematic, as there were few ways for small CCGs to negotiate with the larger organisations, meaning that they could sometimes be at a disadvantage:

Sometimes you feel that actually it’s the tail wagging the dog really rather than us as commissioners going out to providers and saying ‘We want you to provide this service’.

CCG C, interview 1

The awareness of this power differential was enhanced by uncertainty around the motives of secondary care providers, and their desire to truly work in partnership with the CCGs:

Don’t go in a hospital. That doesn’t sit with the financial structure of [the secondary care hospital] and you’ve got that contention. So if I was on that side of the fence I’d be doing everything I could to keep patients in longer, get more money . . . there is this tension between primary care and secondary care . . . there’s a lot of doubt and a lot of speculation that secondary care is manipulating to their advantage.

CCG A, interview 9

Socialisation capabilities across different organisational sectors were often undermined as a result of a sense that there may be different priorities or ways of working, which could be detrimental:

Everybody talks the same language, but do they see the outcomes of dealing with those issues the same way? My personal view is I don’t think they do. I think that whilst everybody would agree that keeping somebody supported in their home for longer is a self-evident benefit for a patient, a service user, how that translates into money moving round the system is a different and knottier issue that people don’t want to face up to.

CCG B, interview 2

However, as mentioned previously, individuals from a common professional background may often share common ideals as an aspect of their professional identity. Within our CCGs, despite the managerial concern that secondary care providers and CCGs may have different priorities in terms of funding outcomes, GPs in CCGs and clinicians in other organisations may work effectively together because of their professional background. This was acknowledged by a public health doctor discussing his relationship with medical colleagues in other organisations:

We do have the same broad shared values because we’ve been through the same training, we’ve been through the same processes, we understand where each other are coming from. Even if we’re not totally lined up I think that we are much more comfortable with each other than is always the case with managerial colleagues.

CCG C, interview 10

Once again, despite the issues that may undermine the socialisation capabilities of CCGs and their wider networks, the engagement of clinicians in the CCG may mediate these problems. Being able to harness this potential contribution is, in turn, related to the development of CCG co-ordination capabilities, as we discuss in the next sections below.

Formal commissioning roles

General practitioners who were engaged in formal roles within the commissioning organisation suggested that their practice-based knowledge about the patient population enabled them to improve commissioning processes. By framing themselves as representing the needs of patients, or the voice of other clinicians, they highlighted their influence over management decisions through the provision of clinical leadership:

I give clinical leadership to try and make sure the decisions we make are made clinically, not just in terms of finances or things, but actually putting patients at the centre of what we are trying to achieve for them.

CCG D, interview 5

Describing themselves as clinical leaders, the GPs drew on their professional standing to influence commissioning decisions. In particular, their expert claims enabled them to elevate themselves over general managers, who were unable to draw on similar practice-based knowledge to enhance their influence:

The commissioners are not clinical and they may think something’s a good idea, but actually in practice having that clinical background just helps you know that it would be useful or it wouldn’t be useful. So that’s a more important insight.

CCG K, interview 14

It was acknowledged by general managers that GP involvement was an important element within the commissioning organisation. The general managers indicated that the expert knowledge held by GPs about their patient population was one of the main influences on strategic decision-making for clinical services.

In particular, formalised GP involvement acted as a co-ordination capability as GPs drew on tacit, or professional, knowledge and experience to enhance critical review capacity. As outlined previously, standardised data sets or systems of reporting may not represent a full picture from which to make commissioning decisions. By acquiring tacit knowledge from GPs, assimilation processes were enhanced as standardised data could be contextualised with professional understanding, overcoming limitations of systems capabilities:

What we’re currently finding is that CCGs often tend to pick on one particular piece of evidence, primarily data, that seems to drive decisions more than anything else . . . the commissioners are pretty good at using data and they love numbers and they love looking at activity and costs, but where the gaps are more around the research evidence, so what we’re saying to commissioners is ‘It’s all very well looking at the data, that’s fine, but we need you to have a much richer, rounder picture of what’s out there to inform your decisions’ . . . that’s where the GPs come in.

CCG L, interview 15

For example, the involvement of GPs as a co-ordination capability was particularly highlighted in examples in which standardised acquisition systems, such as risk profiling tools for patients at risk of admission, were not perceived as comprehensive. Involving GPs in further identification of ‘at risk’ patients subsequently enhanced the scope of knowledge available to guide decisions:

My experience is that the people that it [existing standardised systems] throw up are not all of the people that we need to discuss. So the GPs will bring up other people that haven’t been thrown up by the risk profiling system which they know are ongoing cases that we’re all involved with and we know are possibly at more risk of going into hospital than others. They don’t come up in the system but the GP knows about them.

CCG G, interview 3

Further to their potential as a co-ordination capability to mediate systems limitations, GPs in formal commissioning roles also acted as a conduit between non-commissioning GPs and the commissioning organisation. This conduit role facilitated the flow of information both to and from rank and file GPs. Suggesting that they were able to influence their peers, commissioning GPs took ownership of the profession, referring to ‘my GPs’ and setting themselves apart as leaders of the group:

I’ll e-mail my GP constituents about what’s coming up on the agenda, particularly the things I think they’ll want to make their views known on . . . I’ll say ‘Well actually, I’ve spoken to or communicated with my GPs and what they would like to make known is that they don’t think it’ll work for this reason’, or ‘They think more funding should be given to that’.

CCG B, interview 3

Commissioning GPs consequently worked to mediate barriers of socialisation capabilities, working to communicate clinician opinion to the commissioners, but also attempting to facilitate the uptake of commissioning decisions in practice, to reduce the potential of professional resistance. Through their conduit role, they used their professional background to influence other GPs to align with commissioning priorities:

I try and explain to them about the cost involved and the reasons we should change, and GPs want to be in line with their peers so they usually make the change if you explain it peer to peer.

CCG F, interview 7

Furthermore, in addition to enhancing critical review processes by engaging their clinical peers in the design and delivery of services, GP commissioners were able to overcome the barriers of socialisation capabilities evident in interactions with other stakeholders such as secondary care. Negotiations with secondary care partners were noted by general managers as potentially inhibitory during the negotiations of services:

For me to go in as a manager and try and argue a case with a dozen clinical directors, with the best motivation all I can do is argue the numbers, the philosophy, present a management argument to why we should do this or we should do that . . . they don’t see that as credible.

CCG B, interview 3

In these circumstances, the influence of power differentials between professional groups, representing socialisation capabilities, could be inhibitory. CCGs are commonly small organisations that are seen as less influential than hospitals, the latter being dominated by powerful groups of doctors who have been accustomed to patterns of resource allocation around which they resist any change, whatever the ‘evidence’ might suggest. However, those interviewed once again noted that the involvement of GPs had the potential to mediate limitations of socialisation capabilities. The involvement of GPs acted as a co-ordination capability by encouraging integrated working with clinicians in secondary care organisations and involving them in service design discussions:

When we involved GPs, we saw negotiation moved away quite dramatically from the old-style negotiation which was all about finance and activity to a discussion that focused on quality outcomes and patient pathways . . . the GPs were able to bring a level of reasonableness into that room with their medical colleagues that had previously not been there, with hospital doctors viewing us managers with some suspicion. GPs brought in the perspective of a practitioner dealing with patients on a day-to-day basis, which hospital doctors accepted and which really altered the dynamic in the room.

CCG B, interview 10

However, although GPs in formal commissioning roles suggested that they were able to influence both commissioners and professional colleagues because of their professional background, there were indications that co-option into a formal managerial role had the potential to undermine influence with their rank and file peers. Reducing clinical time to accommodate the demands of their commissioning role led to concerns about their credibility with peers:

Once you become a full-time commissioner, or say just do 1 day a week, you start to lose a bit of your clinical credibility . . . People say ‘Well he only does a day a week, what does he know?’.

CCG A, interview 4

Reflecting concerns that colleagues may perceive them as ‘bloody managers’ (CCG B, interview 1), comments from non-commissioning GPs suggested that the close association with general management of formal commissioning GPs undermined their influence within the profession, subsequently reducing their potential to encourage uptake of commissioning decisions:

Some of them don’t really know what it’s like working day in day out. They go and sit at a desk all day and say they’re representing me, but I’m the one seeing patients all day . . . I don’t care that the CCG has decided this or decided that, it makes little difference to me.

CCG I, interview 9

The extent to which GPs in commissioning roles had continuing influence among their non-commissioning colleagues was thus questionable. Although they could act as a co-ordination capability to some extent, presenting themselves as professional leaders who represent the voice of the patients and of other clinicians, responses from non-commissioning GPs suggested that they were not fully able to overcome socialisation barriers. Commissioning GPs suggested that their influence lay in their ability to act in a formal two-way conduit role, to communicate both to and from the commissioning organisation, influencing both managers and practitioners. However, although their claims of influencing strategy within the commissioning organisation were reflected by comments from general managers, their association with managerial structures and priorities had the potential to undermine their claims as professional leaders, reducing influence among non-commissioning peers.

Practice-based general practitioners in informal roles

In addition to GPs taking on a formal role within commissioning organisations, we identified a group that was actively engaged with commissioning organisations, but did not occupy a formal commissioning role. These GPs worked full time in practice, but frequently communicated information to the commissioning organisation through informal structures:

When I want to voice my opinion about something I usually go round [the formal GP commissioner] and just ring someone higher up, the chief operating officer or the head of urgent care commissioning or whoever it may be. They know me, I’m fairly well-known in the area, and I tell them what I think the issues are and what they should be doing.

CCG F, interview 14

In other words, there was another group of GPs attempting to establish themselves as professional leaders. They presented the front-line, real-time experience of the patient population to managers in the commissioning organisation, in order to influence commissioning decisions:

Rather than the very high level it’s actually ‘What does this actually mean in practice?’ and sort of brings a flavour of the situations that I encounter on a daily basis . . . that voice saying ‘Actually is that really going to work? Is that really how it happens in GP practice?’.

CCG D, interview 3

General practitioners who were engaging informally with the CCG attempted to position themselves as professional leaders, suggesting that they had more influence within the profession than formal commissioning GPs, as those working in formalised roles no longer represented the reality of rank and file professionals:

All the clinical commissioning leads are really well embedded into their roles and sometimes you need somebody to be a critical friend saying ‘Actually is that really what happens?’ . . . You just need to be that person to say ‘But is that really the right thing to be doing?’.

CCG D, interview 10

As such, informal GPs were better placed to overcome the challenges of socialisation capabilities faced by formal commissioning GPs, as outlined above. However, although those informally engaged aggrandised their position by highlighting their distance from the formal organisational management, their influence over managerial decision-making was subsequently reduced. As they did not occupy formal positions, they were unable to attend meetings in which strategic decisions were made, and their contribution was limited to the informal communications with commissioners outlined previously. Conversely, formal commissioning GPs were able to attend formal meetings as part of their role. Consequently, GPs who were engaged informally in commissioning reported less influence over managerial strategies than they might have liked:

I wouldn’t say I feel hugely influential . . . I speak to the commissioners fairly regularly, they ask my opinion and I don’t hesitate to ring them if I think they need to be approaching things differently. But I think the difficult part is not being able to attend meetings . . . I can’t be that person sitting there going ‘What about this?’ when decisions are made.

CCG G, interview 6

Although these GPs were able to communicate their ideas to the commissioning organisation, and were actively engaged in doing so, their influence over managerial decision-making was undermined by their lack of a formal role. However, despite their lack of influence within the commissioning organisation, informally engaged GPs suggested that they enjoyed having more influence among their professional peers. To this end, they positioned themselves as professional leaders among rank and file GPs, communicating the opinions of the professional community to commissioners. Further to this, as a result of their high levels of credibility because of their full-time practice-based roles, they were able to encourage and influence their peers to engage with commissioning decisions:

I think the fact that I’m a full-time clinician helps from a couple of aspects. I think it helps practices understand that there is value to these decisions, but also that clinicians just approach things in a slightly different way. I have more sway than the commissioners, my colleagues listen to me.

CCG A, interview 16

Their full-time position in practice, combined with their informal relationship with the commissioning organisation, afforded them more influence with their peers than formal commissioning GPs:

At the end of the day I’m going to be the one actually using these services every day, or being asked to do whatever they’ve decided. So I have influence in practice which is relevant, and I communicate back to the commissioners about how those things are working, or how they are viewed by other GPs.

CCG C, interview 2

Indeed, during interviews, in contrast to the self-aggrandising narrative of the formal commissioning GPs to communicate their professional influence, it was their rank and file peers who identified the practice-based GPs as leaders of the profession:

One of the partners here is very engaged with the CCG, he doesn’t work there but he’s always e-mailing them, phoning them, pinning them down about things . . . I trust him more to tell them how it is, not to worry about whether it fits with their priorities. I’m sure the others [formal hybrids] do that as well at the beginning, but they must naturally get sucked into the CCG mindset.

CCG H, interview 8

The comments from rank and file GPs suggest that, rather than identifying themselves as professional leaders, those who engaged informally with the CCG were seen by their colleagues as professional leaders, and were more successful in overcoming limitations of socialisation capabilities. Although their lack of formal managerial involvement detracted from their ability to directly influence commissioning decisions, it facilitated their influence within their professional group.

Complementing roles to protect professional influence

In the sections above, we outlined variations in influence in a non-homogeneous professional group. Differentiating between formal and informal commissioning roles, the findings indicated that neither group had influence across both managerial and professional groups. GPs in formal roles acted primarily to overcome systems limitations, drawing on tacit knowledge to enhance commissioning decisions. However, their co-option into managerial roles undermined their influence among non-hybrid peers. For GPs engaging in informal roles, a lack of co-option restricted their ability to influence decision-making processes in the commissioning organisation. However, they were more able to act in a co-ordination capability to overcome socialisation capabilities by influencing their non-commissioning peers.

Exploring how hybrids mediated the limitations of their influence in different spheres, both formal and informal GPs indicated that they worked together to ensure that there was continued professional influence over managerial decisions. For example, an informal commissioning GP noted how a balance between formal and informal roles maintained a focus on the needs of patients:

There are a few clinicians who have devoted much more to this, but then they’re not doing very much clinical practice at all . . . but I think hopefully in our area we’ve actually quite a good balance between those who are dedicated to it full time and those who come in on a more limited, informal basis like myself. And we work together for the good of the patients.

CCG H, interview 8

Both groups of GPs discussed episodes during which they worked together to enhance or protect the status and influence of the profession as a collective. During interviews, both groups noted examples of when they had worked together, drawing on either the managerial influence of formal GPs or the professional influence of those in informal roles. For example, one formal commissioning GP highlighted how they were able to use their understanding of managerial structures within the organisation to maximise the influence of the profession as a collective:

I know how the system works. I think I’ve used that to my benefit, to the benefit of colleagues and their patients by knowing who to go to when something doesn’t work . . . And so now the GPs might phone me and I know that if I make a couple of phone calls I can escalate that to the right contracts manager who can then put pressure on the provider to make those services easier [to access]. So there are lots of things like that where me having the commissioning knowledge and my colleagues in practice can come together to do good things.

CCG A, interview 4

Similarly, GPs engaging informally noted how they were able to act as mid-level brokers or mediators between the commissioners and their rank and file peers. In particular, one informal GP who enjoyed high levels of influence within their professional community, and who regularly engaged with the commissioning organisation, was able to communicate information and feedback across professional and managerial boundaries during a service dispute between GPs and the managerial staff at the organisation:

I knew the [commissioner] to talk to, so I had a word with them and made it clear what the GPs were concerned about, what they could do, and then I fed that back to my colleagues . . . at the moment I feel I’ve done my initial bit which is be the middle man and mediator with it . . . I’m not tied into the CCG, so I feel a bit more of a free agent to move between the two groups. I actively stay out of the formal commissioning part, and that means I’m not seen as such by my colleagues.

CCG D, interview 10

The GP in this case highlighted how they actively chose to remain in full-time practice, rather than taking on a formal hybrid role. In doing so, they were able to achieve influence by selectively engaging with managerial structures and commissioners without being co-opted.

Although the two groups of GPs used their professional or managerial influence in different ways, they were aware of the need to remain as two distinct groups, rather than diluting the formal or informal nature of their role to make the group more homogeneous. Both formal and informal roles were seen as fundamental to the ability of the profession as a collective to influence service change:

I take the view that my [GP] colleagues want to be left alone to get on and do their day job . . . I don’t expect all of them to be involved in service redesign because not everyone wants to do that . . . So I don’t expect them to be experts on that and how could they be? That’s a job that I take on to do for the CCG. And I’ll work closely with my peers who are influential in practice and that’s how we can achieve change.

CCG G, interview 10

At the same time, all of those who were interviewed suggested that their professional collegiality, and commitment to buffering professionalism from managerial influence, encouraged both groups of GPs to work together to retain professional control:

I feel that now, if we work together, we do have a genuine chance to possibly change the way the system works.

CCG A, interview 4

By working together in formal and informal roles, GPs could use their influence to achieve change within the commissioning organisation. For example, one informally engaged GP discussed a new service that had been designed and set up in collaboration with commissioning peers, and the resulting positive feedback from rank and file GPs:

Initially there was a lot of scepticism . . . But I worked very closely with [formal GP commissioner] and they pulled strings in the CCG . . . and I worked very closely with my colleagues in practice, so now I’ve had a lot of my colleagues say ‘well, it actually sounds like you’re doing something rather than just sitting round a table and talking endlessly about things and not actually making any changes.’ So we’re now actually in the roll-out of that project and both the commissioners and the GPs are on board with it.

CCG K, interview 5

Ultimately, through collaboration, the two groups co-opted managerial processes, drawing on peer influence to encourage commissioning decisions that were aligned with the values of the collective profession, rather than driven by managerial demands. Although formal GPs actively took on commissioning roles, they combined their managerial knowledge of the commissioning organisation with the practical experience and influence of those in informal roles to enhance professional practices and protect professional jurisdiction. In doing so, they buffered the professional group from control by general commissioning managers.

In conclusion, by working together, GPs in both formal and informal roles were able to co-optate managerial processes and maintain a form of professional resistance to managerial infringement.

Conclusions

General practitioner involvement acted as a co-ordination capability in two ways. First, by drawing on tacit or professional knowledge, GPs could enhance critical review capacity by overcoming systems limitations. Second, they engaged in clinical leadership to mediate socialisation limitations by using their professional background to influence peers in both primary and secondary care. This enhanced service design and delivery, and reduced the potential for professional resistance. There were variations in the way GPs engaged with the CCG, as some took on formal roles and others engaged more informally. Although formal GP commissioners had more influence over commissioning decisions, they may have limited capacity to overcome socialisation capabilities, because of their proximity to formal managerial roles. Those in informal roles had more influence over their peers, but faced inherent challenges as a result of the time limitations of their involvement and maintaining a clinical practice. However, we also noted how these two groups worked together to protect professional jurisdiction and ensure continued influence over commissioning decisions. We highlight the importance of this non-homogeneous professional group, and do not suggest that all GPs need to become formally engaged in CCGs, as their clinical practice is an important element of their professional influence. Commissioning organisations should be aware of the need to balance formal and informal engagement with GPs to maximise their potential as a co-ordination capability.

Clinical Commissioning Group D: rational ideologies of managerial control

The following extract is from field notes taken at the second PPI group meeting attended by the research team:

The group was asked by the lay chair to identify what they feel their contribution to the CCG should represent. There was a lot of confusion amongst members about their potential contribution . . . The chair had difficulty keeping the group on point . . . Two general managers attended the meeting, and they were responsible for setting the agenda. By the end of the meeting the managers had taken over the running of the session to ensure all the agenda items were discussed in a more structured manner.

Field notes, 11 February 2014

This observation of the PPI group at this site highlights two influences on PPI: (1) confusion among public representatives about their role and potential contribution to the organisation and (2) the dominance of managerial control over the group. The management team took over the running of the PPI meeting as the group was not seen to be conforming to the agenda predetermined by the managers. When interviewed about the PPI group, a general manager said:

It is hard work . . . We’ve spent countless hours doing terms of reference, loads of time explaining what their role is . . . they don’t necessarily understand how the NHS works so that’s part of the problem . . . people’s understanding of the health system is actually really quite minimal and they don’t understand all the interorganisational arrangements.

General manager, interview 5

The perception on behalf of the management team, reflected above, seemed to be that the PPI group was not informed enough about the NHS to contribute meaningfully about decisions made by the commissioning organisation. Indeed, the manager notes how they spent time ‘explaining what their role is’ to the group, limiting PPI contribution to a managerially determined role, rather than encouraging the group to determine a role for themselves. This represents a limitation of socialisation capabilities. In addition, the perception that they had ‘minimal’ understanding of the issues in the health service often led to derisive comments from managers during discussions about PPI:

They’ll randomly e-mail me going ‘oh [manager’s name], have you seen this thing about telehealth?’ and I’m like ‘oh god.’ . . . they don’t understand the priorities or the organisation, so it’s really difficult to manage them.

General manager, interview 7

Again, the suggestion above is that the PPI group needs to be managed. Managers, therefore, did not accommodate PPI contributions outside those considered to be in line with managerially defined organisational priorities. The influence of the hierarchical organisational context in encouraging rational ideologies of control was illuminated by many of the managers interviewed, who suggested that they often limited PPI to smaller, peripheral discussions, rather than involvement in strategic decision-making. Consequently, there was relative diminishing potential for the PPI group to influence decision-making processes and act as a co-ordination capability to enhance critical review capacity. This sentiment was also voiced by doctors involved in the commissioning organisation, suggesting that the hierarchical organisational context encouraged rational ideologies of control across professional groups as well as managers:

I think them feeling involved is probably the best that we can do. They don’t necessarily have all the information or the knowledge and experience to make the decisions that we would make as health professionals and they do make some really valid points but it’s really peripheral stuff to be honest.

General practitioner, interview 9

Consequently, PPI involvement in decision-making processes was marginalised, and the management team characterised PPI in a way that suggested that they did not see much strategic value in any PPI contribution:

They want better plates or something ridiculous . . . But then again, you see, if you’re a member of the public and you know nothing, like a lot of them don’t really about how the NHS works, then why shouldn’t they bring these things up? It’s not really that helpful for us though, it doesn’t change the way we work.

General manager, interview 3

As noted by the manager interviewed above, PPI contributions were not considered to be changing organisational processes. During interviews, none of the managers or PPI representatives could give examples of the way in which ideas from the PPI groups influenced development of services, demonstrating that, in this case, PPI was not working as a co-ordination capability. In addition, there was little capacity for new ideas to be generated from the PPI groups, as the management team set the agenda, constraining PPI contributions to managerially determined issues. As such, the potential for new, or different, ideas to emerge from the PPI group was limited:

There aren’t many things that come up that don’t really fit with the main themes because we set the agenda to make sure that what they’re discussing fits with the main themes.

General manager, interview 3

On the one hand, organisational managers were aware of the need to engage in PPI in line with national policy. However, their interpretation of policy led them to engage in rational ideologies of control, which constrained the PPI group, allowing them to contribute only through formally determined, and managerially led, discussions. For many PPI representatives, the lack of flexibility for their involvement led to feelings of frustration and questions about their potential for strategic influence:

It’s difficult to see how you make a difference really. I think a lot of the time the managers just sit there and nod their heads, they don’t hear us. And we’re not supposed to talk about a lot of things, just stick to the agenda or you get into trouble!

PPI representative, interview 10

However, the chairperson of the group, who also represented the PPI group as a lay representative of the commissioning board, described how he felt that he was involved in strategic-level discussions. An experienced manager in the private sector, he had been appointed as chairperson of the group by commissioning managers, who identified him as ‘our champion for public and patient involvement’ (general manager: interview 1). As such, the chairperson felt that he could use his professional background, and the influence afforded to him by his managerially appointed position, to contribute to managerial decisions:

I have some skills from a professional management point of view . . . So as a non-exec[utive] sitting on the board I’m able to say ‘Well, why do you do it that way? Have you thought of . . . ?’ So I’m able to put constructive challenge into the system.

Lay member, interview 2

Therefore, by drawing on previous experience of management within different organisational contexts, the PPI representative achieved more strategic influence by working within a managerially framed role. However, he was also aware that, to retain this influence and his position given to him by managers, he had to behave in a way that aligned with managerial interests or accepted behaviours, influencing the way he communicated information from other PPI representatives to commissioning managers. In this way, he could be seen to replicate forms of rational control over his PPI group members, replicating the barriers of socialisation capabilities:

If I go to that board with a proposal which is unrealistic, then I’ve failed . . . I have to go back to the [PPI group] and say is ‘It’s all very well coming up with these ideas. How do we fund it?’ They have a responsibility as well. They can’t just say ‘We want a cancer consultant on our doorstep every day of the week’ like sometimes they do . . . So if I take it into the board I have to make sure it is credible. If it’s not, then I’m going to get thrown out and rightly so. I’ll have a fight and I won’t be taken seriously.

Lay member, interview 2

CCG D represents an organisational context in which the potential of PPI to act as a co-ordination capability is undermined by socialisation capabilities. Specifically, rational ideologies of managerial control undermined opportunities for PPI. An organisational context, which did not prioritise the potential for PPI outside formal structures, meant that managers set the agenda to limit PPI contributions to peripheral issues, rather than allowing strategic decision-making processes. Consequently, there were no examples of the PPI group influencing the commissioning organisation at a strategic level. Although the managerially appointed chairperson of the group felt that he had influence, this was related to his ability to work in a way that was consistent with managerial expectations, as a result of his professional background and previous experience, and having a position bestowed on him by the commissioning managers. The need to act in a way that was aligned with managerial interest, through channels determined by rational ideologies of control, subsequently influenced the type of information the lay member communicated between the PPI group and the management team, as the chairperson himself replicated forms of managerial dominance and reinforced socialisation barriers.

Clinical Commissioning Group G: increasing recognition of the potential of public involvement

In CCG G, characterising their attitude towards PPI as a ‘supportive culture’, managers were keen to communicate the importance of PPI within the organisation, highlighting that their organisational context encouraged and recognised potential benefits of PPI in service design and delivery:

If we actually end up designing services or commissioning services that have got strong, local backing it’s going to make it 10 times easier to actually put those services in place because you can say ‘well actually, we’ve got the backing of the general public here and this has all been informed.’. Yes, we’ve been influenced by our clinicians and the GPs, but particularly the public.

General manager, interview 7

However, although all managers claimed that they valued PPI, some voiced concerns over the level of understanding of PPI representatives about the commissioning organisation, and the context of the wider health system. Similar to CCG D, some managers acknowledged that a lack of awareness of the wider organisational landscape could limit PPI:

What’s practically quite hard is having to go over things that we’ve already talked about, and so something maybe you talked about 3 weeks ago you then spend half the meeting going through it again. It’s quite time-consuming and can be quite frustrating.

General manager, interview 13

Despite what was characterised as a more supportive managerial culture, managers once again resorted to rational ideologies of control, and limited PPI to topics or contributions determined by the management team. However, the chief executive of this organisation suggested that a need for control was not because of a reluctance to involve the public in strategic decisions, but was about identifying the most relevant opportunities for involvement, arguably enhancing their contribution towards decision-making:

There is always a line to be drawn between what is operational detail that we sort out every day, kind of day-to-day stuff, and what is really important, relevant stuff that the public ought to know about . . . So sometimes it’s not conducive to ask them about it.

Chief executive, interview 11

Although managers acknowledged that their dominance could limit the potential of PPI as a co-ordination capability, trying to focus the contribution by involving PPI representatives in ‘relevant’ discussions, they suggested their organisational culture was one that valued PPI. Diverging from the findings in CCG D, in which organisational approaches to control trivialised PPI and led to concerns from PPI representatives that they had no real involvement, PPI representatives in CCG G seemed more satisfied with their contribution. In particular, they suggested that they were able to challenge managerial decisions and overcome limitations of systems capabilities by raising issues outside the managerially determined agenda:

I think they’ve realised that we are a real asset so, you know, I think we’ve got a good working relationship . . . they’re very welcoming and, you know, we can be very challenging and raise some really difficult questions, but then they do try and seek the answers. We can ask about anything really.

PPI representative, interview 4

However, despite positive attitudes towards PPI in CCG G, which might have engendered more normative forms of control, the way in which PPI influenced structure and recruitment to the PPI group was tightly controlled by commissioning managers and, as such, represents a further limitation of socialisation capabilities. PPI representatives were not able to offer contributions to the commissioners outside the boundaries of managerially controlled mechanisms for involvement:

I was appointed by the commissioners to the committee . . . So the public interact with [PPI] committee; the [PPI] committee collates the information and then we transmit this in one way or another to the clinical commissioning group.

PPI representative, interview 2

This was particularly frustrating for one PPI representative, who acknowledged that the organisational context was supportive of PPI when occurring within controlled strictures, but questioned the extent to which he could contribute to wider issues:

They didn’t discourage me from doing research, but they couldn’t say if a layman like myself could carry out something . . . so now as an active member, quite active and outspoken, I’m still feeling somewhat tagged on . . . Are we really important or is it all window dressing? I’m allowed to have an opinion but only on topics they allow me to. So I just wonder what my contribution is.

PPI representative, interview 1

Therefore, although more anecdotally committed to PPI than CCG D, PPI was still influenced by socialisation capabilities, limiting involvement to a small number of preselected topics. Although some PPI representatives were satisfied with their involvement, perceiving the organisational context as one that supported and valued PPI, others felt limited in their contribution. For this reason, although the managers in this organisation attempted to enhance PPI, facilitating involvement where it could have the most impact, inhibitory socialisation capabilities in the form of managerial dominance had the potential to exclude PPI representatives from involvement in more strategic issues. This led to some members feeling ‘tagged on’ and not integral to the commissioning process, and undermined their potential as a co-ordination capability.

Clinical Commissioning Group C: the potential of normative control

Clinical Commissioning Group C is notable because of the different approach to the implementation of PPI policy taken by the management team. At this site, the managerial culture encouraged a more normative approach to the organisation and control of the PPI group. Managers were significantly less controlling in a formal sense and described themselves as holding a greater commitment towards the diverse viewpoints that PPI potentially brought into decision-making. They thus claimed a ‘hands-off’ approach to PPI.

The management team was not actively involved in the development of agenda or running of meetings, and the PPI group had a small budget provided by the commissioning organisation, which they could use for their own purposes (such as room bookings, transport and small research projects). In addition, the PPI group members elected a chairperson and determined the group recruitment strategy independently of managerial control:

We decided to try and have a representative body and my original thought was that we’d have a group with one representative from each of the patient participation groups and any special interest groups would bring in a representative . . . The other thing we did, we decided that if people showed a strong enough interest then we’d take on a few independent members.

PPI representative, interview 2

Despite the apparent lack of direct involvement from the commissioning managers, the PPI representatives reported excellent relationships with management. Some representatives had been involved under previous organisational structures and noted the difference under the new commissioning arrangements:

[Previous organisations] were very much management led . . . they tended to do things on business lines rather than on patient-need lines . . . everything was organised because that seemed like the right way to make it easy for somebody who was organising it rather than ‘what actually are we trying to get out of this? What does the patient really need?’ . . . So I have seen a major shift.

PPI representative, interview 1

Rather than determining the topics the PPI group should be involved in, the management staff allowed the group to discuss issues they felt were important. This encouraged the generation of new ideas for service development:

We don’t tell the PPI group what to discuss, they come up with the ideas and set their own agenda, arrange their own meetings. And sometimes they’re used as a sounding board, so they will often come up with ideas, very good ideas which we will use in commissioning design.

General manager, interview 7

During interviews, managers noted that their normative approaches to control – claiming that they set aside organisational hierarchy and encouraged a collaborative ethos – not only enhanced the PPI contribution, but developed a sense of organisational belonging or ownership among PPI representatives. This subsequently influenced the implementation of the services in a wider public setting:

If you use patients to help in the design you may still end up with the same result, but because you’ve taken patients along with you they will go out and defend it themselves . . . at a public meeting somebody was having a go at our CCG over something and one of the guys from the patient group stood up and defended it because he’d been in on the inside and said ‘you’ve totally got this wrong. They’re doing it this way.’. So it shows it does work.

General manager, interview 13

In contrast to CCGs D and G, both managers and PPI representatives interviewed in this case could give explicit examples of episodes when PPI had influenced service design or had been clearly involved in strategic decision-making processes. In other words, PPI, in this case, acted as a co-ordination capability to overcome limitations of socialisation capabilities, enhancing the critical review capacity of the CCG. As noted above, this also encouraged a sense of organisational commitment and ownership of services, and instilled a feeling among PPI representatives that they were able to contribute at all levels of commissioning processes. This was reflected in an example from one of the PPI representatives who was concerned with the local ambulance service and, as a result, undertook their own research, independent of the commissioning organisation. After developing a paper and arguing a case for change in the way services were being delivered, the PPI representative distributed it to the PPI group, the board of the commissioning organisation and the board of the local ambulance service:

I sent a paper off to the chief executives because I’ve been concerned about the lack of data transparency from the ambulance service . . . So I wrote a paper and it’s very challenging potentially . . . I’ve pursued it, asked questions.

PPI representative, interview 10

The above example illustrates how a more collaborative ethos towards PPI encouraged representatives to become more actively involved in strategic issues outside formalised, managerially determined roles, overcoming socialisation limitations. As a result, there evolve ideas or opinions that may not otherwise have been facilitated, enhancing the critical review capacity of the CCG. A second example of the positive influence of PPI was illustrated in the design for a respiratory [chronic obstructive pulmonary disorder (COPD)] service:

One of the chief executives came to me and said they’d been looking at the plans I’d sent them for a project on COPD, and they wanted my input in developing the service . . . it’s now showing something like a 17% reduction in unplanned admissions. Well, if you add up £3000 or £4000 a time that’s a lot of money we’re saving . . . So the patients are benefiting and the CCG are benefiting because the patients are benefiting.

PPI representative, interview 4

By having the confidence (and managerial encouragement) to send ideas or project proposals that may not be managerial priorities to the management team of the organisation, PPI encouraged the development of a new service, subsequently resulting in patient benefit and cost savings. If the PPI representative, in this case, had been constrained by socialisation capabilities, such as those reflected in CCGs D and G, which were characterised by rational control, it is unlikely that the service would have been developed, as it was not part of the immediate managerial agenda.

Public involvement representatives in CCG C seemed aware that their potential influence within the CCG was not reflective of the types of involvement in other areas. As a result, some members within the PPI group were reluctant to work with other ‘less credible’ PPI groups in case this undermined their influence. This aversion to actions that may reduce their level of influence within the commissioning process was noted in field notes during a discussion at a PPI group meeting about the potential for collaboration with a neighbouring organisation:

[Discussion about collaboration with neighbouring organisation]. Everyone becomes very animated and derogatory about the PPI rep[resentative]s in neighbouring area. Many people appear concerned that the other group have less ‘credibility’ as patient representatives, and that they do not take their role as seriously. A number of people suggest that they are ‘not the same as us’ and ‘don’t know how to work with their CCG in the best way’. Someone says they are just ‘yes men’ and will not challenge managers.

Field notes, 9 November 2013

The PPI group seemed aware that its influence came from its ability to work alongside commissioning managers but remain outside managerially determined roles. Working in an organisation with a more normative approach from managers towards their contribution did not require PPI representatives to work within rationally controlled involvement structures, and afforded them freedom to think more innovatively about service design as they were not constrained to managerial priorities or visions. The distinction with the neighbouring group was that it was seen as being particularly influenced by a more hierarchical structure that privileged managerial priorities:

In [the other organisation] they’re appointed by management . . . The thing is when you suggest something the people from [the other organisation] will put all the obstacles in the way and the people from here will say ‘well why can’t we do it? What’s stopping us?’ . . . You see, I wouldn’t hear ‘we can’t do that’ from us, but I would from them.

PPI representative, interview 6

The PPI representatives were aware of the unique freedom afforded to them by the commissioning management team, enhancing their influence and potential contribution. Their aversion to working with managerially controlled groups was a result of concerns that they may find their influence lessened by increased pressure to conform to managerial priorities. However, the PPI group were also keen to help educate others and encourage them to work independently of managerial agenda. This was highlighted during interviews by an awareness of their role in developing influential PPI representatives for the future:

I’ve been to meetings and I’ve asked questions and I’ve heard somebody say afterwards ‘he shouldn’t have asked that question. He knew the answer.’ Yes, I did know the answer, but half of the people there didn’t which is why I asked the question and they wouldn’t have known to ask the question . . . So that’s where I see my role. I see my role now bringing those people along, helping them along so we’ve got some people who can question things in the future.

PPI representative, interview 10

In conclusion, CCG C illustrates the potential for PPI to act as a co-ordination capability when it is subject to normative, rather than rational, ideologies of control, meaning that it is not constrained by socialisation capabilities, and is capable of realising the aims of PPI policy. By being able to work outside rationally controlled roles, while maintaining supportive relationships with commissioning managers as a result of normative control, PPI representatives could inform commissioning processes at a strategic level and had a clear influence on the development of services. In addition to encouraging organisational commitment, and ownership of services within the wider public, normative control of the PPI group enhanced the ability of the commissioning organisation to design responsive services, resulting in positive outcomes such as the COPD service.

Conclusion

Public involvement is a legal requirement of CCGs and was shown in varying levels across all 13 cases, but notably in three situations. We identified that rational, managerial control inhibits the potential of PPI as a co-ordination capability by restricting involvement to managerially determined topics and areas. Most CCGs were increasingly aware of the need for ‘genuine’ PPI, but socialisation capabilities again undermined the potential of true engagement. One CCG took a more normative approach to PPI, indicating an approach that enhanced the potential of PPI as a co-ordination capability, resulting in positive outcomes in the design and delivery of services.

Embedded business intelligence teams

As noted, embedded BI teams were situated within the CCGs, rather than offering BI services from outside the organisation. One of our cases (CCG E) had integrated its BI unit into the commissioning organisation, rather than contracting with an external CSU, resulting in a BI team that was fully embedded in the CCG. This was rare, and not seen in our other cases. Most commonly, CSUs were contracted to provide both embedded staff, who worked at CCG offices, and services from staff who worked externally to the CCG at CSU headquarters.

In both BI formations, interactions between embedded BI staff and commissioners were relational, rather than contract-based and transactional. The embedded teams were able to draw on social resources to overcome limitations of socialisation capabilities by being seen as ‘one of the team’, and working together with the commissioners towards a shared end goal:

And I thought driving down here today that’s what makes us different, is we are in-house and we are a team from that point of view and we have the same ethos. We’ve built our vision together, we’ve built our strategic core principles together. All of those things actually we are in us. That’s not a service we buy.

CCG E, interview 15

For embedded BI teams, their physical location within CCGs, rather than working remotely, was also seen as facilitating their assimilation processes:

It’s more reactive I think and you can overhear things and be involved in more things if you’re on the ground with the people who are making the decisions. I think if we were a central team rather than an embedded team you’d find you’d just get asked for a number and then they would go away and do it, whereas this is more they’ll ask you for a number but then you can add more understanding to what that number represents.

CCG M, interview 4

The local knowledge of embedded BI teams was a crucial mechanism, which allowed them to act as a co-ordination capability. By drawing on prior relational experience with both commissioners and providers, embedded BI teams were able to overcome some of the limitations of systems capabilities:

You’re getting local knowledge, which is actually priceless when you’re doing the kind of work we’re doing. We’re dealing with large data sets and we know the ins and outs. Because we’ve worked for so many years with our local hospitals we know the ins and outs of their data . . . having the local, in-house knowledge is very valuable. It stops people from going off on a bit of a whim redesigning pathways, which perhaps don’t need to be redesigned because there are no savings there to be had anyway.

CCG E, interview 14

The local knowledge of those working within the BI unit encouraged lateral communication between the BI unit and the commissioners, preventing ‘people from going off on a bit of a whim redesigning pathways’. Embedded BI teams also overcame systems limitations by giving meaning to the ‘pure numbers’, as described in the following quotation. Rather than providing decontextualised information, local BI teams could tailor information to local need, enabling commissioners to analyse and apply information to the situation at hand:

If I was just asking somebody for activity data relating to a particular service who didn’t know the area then I’d probably get exactly what I asked for. If I ask for that here I’ll get additional information and insight around ‘but these used to be coded differently. This time around . . . ’ or ‘there’s been a change in the contract here,’ or ‘when we talk to the provider about this we know they’re coding it in a particular way’, and that really, really helps when we try to describe an implementation or try to work out what the benefit of doing something is. It’s more than the pure numbers. So that’s really good and it helps because it’s individuals who understand the locality and we have a relationship with.

CCG E, interview 16

In particular, the relational nature of interactions between embedded teams and commissioners was beneficial in encouraging the engagement of GPs in commissioning discussions. Developing relational interactions with GPs was seen in some cases as reducing the potential for professional resistance, as embedded BI staff were able to respond to immediate concerns of local clinicians, rather than being viewed as detached or separate from the needs of clinicians:

A lot of the things I’d picked up from being in the GP meetings allowed me to feed into that some of the issues and concerns that GPs were having on a local level and concerns on specific projects that are coming up where the GPs are going ‘well, we’ve tried this four times before and it hasn’t worked. Why are we trying it again?’. So that side of things . . . I like to get as much information as possible to try and add more value to everything that we’re trying to do because it’s very easy to just give somebody a number and say ‘well this is what it is’, and I think you need to have an understanding of why they’re asking the question and what impact it’s going to have on everyone around.

CCG M, interview 5

In addition to a potential reduction in professional resistance, embedded BI teams were better situated to involve clinicians in the assimilation and understanding of standardised data sets. By working closely with the BI team, clinicians could use tacit knowledge to influence service design and provision:

I think it was eye-opening to him what we could do with the data, and how we could build a service out of that, and it was eye-opening to us because I wouldn’t have known ‘golly, if we did that and that we can actually inform better practice.’. And that’s all about if you inform better practice you’re keeping people out of hospital.

CCG E, interview 5

However, although the embedded team was well positioned, as a co-ordination capability, to mediate socialisation barriers, there were concerns that its ability to provide timely, locally appropriate information to commissioners was limited by systems capabilities. In particular, restricted access to patient information was seen as a barrier undermining the embedded BI team’s critical review capacity:

So from April 2013 we were not allowed to see patient-identifiable data any more and the primary care data that we used to be able to access was turned off . . . Now if we can’t extract the primary care data to be able to say who’s [turning up] at the GPs with different long-term conditions and how they’re being managed at the GP surgery and all we’re seeing is the secondary care element of that when they’re going through as an emergency admission, what we haven’t got any idea of is what’s happening to them in primary care, or even then when they step down and go into social care we don’t have any of that data. So that to me is the massive gap that is missing in the data that we’re doing.

CCG M, interview 1

Further to this, the embedded teams voiced concerns that the detail of data they were acquiring, even from secondary care providers, was not comprehensive enough to fully answer questions from their CCGs:

Some of the data that we get is very difficult to scrutinise . . . It’s a very aggregated data set so it’s very difficult to try and actually drill down into that data to see whether you really think there’s anything wrong . . . we can’t drill down.

CCG M, interview 2

A second limitation to the critical review potential of the embedded teams was their capacity to respond to ad hoc, time-sensitive requests for information from commissioners. As noted earlier, the strength of the embedded team lay in their relational, rather than transactional, interactions with the commissioning team. However, this could also be problematic, as embedded BI teams tended to be small and limited in their capacity. As a result, high levels of demand and fragmented requests for information could undermine the critical review capacity of the CCGs:

I do think we have a strategy as a CCG, but our approach to achieving that is a bit scattergun and we get flavours of the months . . . It just feels probably we’re trying to fight too many battles at the same time and perhaps because of that it’s difficult to succeed or understand any of them because you can’t focus on them for long enough. We’ve become the bottleneck really and it’s an uncomfortable position to be in when you know you want to support them all . . . we’ve got no kind of defence mechanism or control mechanism for workload . . . we’re just sort of at the whim of who needs what and often it’s very urgent.

CCG E, interview 3

In addition, embedded BI units reported little or no influence on exploitation processes. This seemed to be related to the limitations of systems capabilities, in particular the underdeveloped nature of data sources relating to evaluation of services, undermining the ability of the BI unit to integrate evaluation data into ongoing commissioning decisions. A particular challenge related to how they might measure the effect of absence of the service intervention they had commissioned to reduce avoidable admissions of older people into hospital:

I think some of the difficulty is capturing what doesn’t happen. Sometimes you will put in a service to prevent avoidable hospital admissions or perhaps deterioration in the health of an older person that might eventually lead to admission. However, there’s a real dilemma about how you evidence the impact of the new service, what would have happened if it wasn’t there. Our data management systems aren’t sophisticated enough for this.

CCG G, interview 2

Finally, it was of concern, even where BI was in-house, when commissioners were asked the question about evaluation of interventions designed to reduce unnecessary admissions of older people to hospital, that they admitted an absence of such a feedback loop:

So how is the virtual older persons’ wards intervention working to reduce admissions?

We don’t gather data about that. Maybe we should evaluate though since I guess we don’t know whether to continue or decommission it.

CCG I, interview 2

In conclusion, embedded BI units are able to draw on relational interactions to overcome socialisation capabilities, and incorporate local or tacit knowledge to mediate the barriers of systems capabilities. Thus, the embedded BI unit can be conceptualised as a co-ordination capability of the CCG. The unit is embedded in the organisation and can contextualise the information acquired. The unit has the right people generating the right knowledge at the right time. However, critical review capacity could be enhanced if the exploitation of information could be improved. More research is needed into the most effective way to enhance this co-ordination capability with a view to improving exploitation of information and knowledge, and subsequent improvement of critical review capacity. In addition, it is important to note that the level of embeddedness varied among CCGs. Some units operated purely in-house, whereas others were employed by external CSUs but physically worked within CCG headquarters. Therefore, there may be national variation depending on CCG structure.

External business intelligence teams

Centralised BI teams were external organisations to the CCGs, denoted as CSUs. Offering BI based on more of a consultancy model, CCGs bought services from CSU teams to varying degrees. At a basic level, external BI could offer similar services to those seen with the embedded teams. However, their potential as a co-ordination capability lay in their ability to offer a more systems-level overview in response to information requests. This was facilitated by their ability to view data from across a regional area, rather than a local CCG area:

The benefit obviously of being a CSU is that we’ve got access to everybody’s data . . . You know, if you’ve got an embedded analyst that’s employed by the CCG they can obviously only see their own data, the CCG data. The benefit with the business intelligence department here is that we can benchmark. We’ve got full access to all the secondary care data for their own CCG plus these 23 other CCGs that allows us to be able to point out where they’re very different or where stuff’s costing more than what other CCGs are paying.

CCG M, interview 1

External BI units could subsequently enhance critical review capacity by linking multiple data sets to provide cross-comparisons of assimilated data from different CCGs across a regional area:

So it’ll be can we do some benchmarking and have a look at our ACS [Adults and Community Services] conditions for our over-75s, for instance. How do we compare to other CCGs? How do we compare over time? Are we getting better or are we getting worse? What would be the benefit to the CCG if we managed to reduce our admission rates compared to the best in our patch or whatever? Things like that trying to quantify potential cost savings and potential admission reductions.

CCG M, interview 2

By using multiple information sources to contextualise the acquired information, the external BI units facilitated the generation of a systems-level overview for CCG performance. This provided CCGs with a more comprehensive ‘picture’ of the flow of patients throughout the entire system:

You can actually follow whole person pathways. So you can look at how people come into the system, how they leave the system and each of the points along that . . . and then there’s an understanding of the demographic characteristics of those, so what’s the age, gender, some kind of social marketing data, characteristics of these cost groups and why they’re different whilst looking at the services that make up those groups . . . So we’ve been creating baselines for people so they understand how people kind of work within that system . . . So if you looked at reducing things like delayed discharge in hospital you can actually see where those people go to now, so you might be able to model whether that would have an impact on things like intermediate care from the community, re-ablement from social care and whether it actually has a perverse effect and increases permanent admissions to residential nursing homes.

CCG M, interview 10

An integrated perspective, overcoming systems limitations, was particularly important when considering the interplay between health and social care, the data systems of which are disjointed:

There’s a lot of activity going on out there and there’s a lot of disjointed information systems I suppose across the NHS, so for me it’d be picking all those flows up, centralising those flows and then working with CCGs to say what’s really going on, what’s really happening to patients around the system . . . So it’s having the capacity to pick that up and look at it as a holistic view from a pathway angle and then obviously help to commission better services for patients.

CCG M, interview 5

In addition, although embedded teams suggested that systems limitations could undermine their potential to act as a co-ordination capability, the external BI team noted fewer limitations. To overcome the barriers to using data systems, the external team acquired information from more diverse sources or used the data in more creative ways:

I know people often feel that absence of data or poor data quality are obstacles, that they can’t address the particular question they’re interested in. My perspective has always been slightly different. I just think that you’ve got a question, you’ve got to make that decision. You deploy whatever data you’ve possibly got at that time in whatever form you’ve got it to answer that question as best you can. There’s no point just shrugging and saying you can’t do anything . . . we have to go with what we’ve got really and part of our job is knowing how to make best use of the data sets that we already have, thinking of different creative ways of using the data to answer the questions they’ve got and knowing which fields or records we should just ignore because we know the data quality’s not sufficiently robust.

CCG M, interview 11

For example, one member of an external BI unit noted how they acquired information from another external organisation to consider the impact of weather on increases in urgent care provision, adding an extra dimension of insight to a question from a CCG:

You know, there’s lots of other stuff out there that’s accessible to anyone that you can just download, so depending on what question we’re going to answer we’ll decide how we might provide evidence for it and then go out to look for the various data sets that we might need. So for example, when we were looking at some of the unexpected increases in urgent care provision, one of the hypotheses was that it had been a particularly bad winter in terms of being really cold that winter or really wet, so we’d go out and perhaps get some data from the Met Office to try and substantiate that claim or disprove it and those sorts of things.

CCG M, interview 8

In particular, the strength of the external teams lay in their ability to provide more in-depth analytical responses to complex questions from CCGs:

So it’s the questions that are much more difficult to answer, so things like ‘what are the impacts of doing X on Y? What’s the likely demand on our services in the future? How has this service done compared to another service in another area?’ or ‘What is the evidence for this service to change?’. It’s kind of those more difficult questions which can’t necessarily be done with routine processing of data and reporting on routine data. We often have to link data and we have to provide a variety of different perspectives on a particular service or clinical area.

CCG M, interview 9

For example, an external BI team member outlined how they were working with one CCG to consider the impacts of service redesign:

They started from the fact that they’ve got a single acute provider working across two sites with basically unsustainable services, in their view, into the long term because of insufficient capacity. So they wanted to run a reconfiguration programme to design a new clinical model for that patch and then out of that clinical model to run all the processes necessary to try to identify a preferred option for reshaping their hospital system.

CCG M, interview 12

However, the external teams noted that their co-ordination capabilities could be undermined by the attitudes of CCGs themselves. They voiced frustrations that the products they provided to CCGs were not being used, limiting their impact on the critical review capacity of the CCG:

I suspect one of the difficulties we have is once we’ve done a piece of work making sure that that work is taken forward and used. Because we don’t work for them or we’re not embedded with them we’ll do a piece of work and then hand it over and then that’s the end of our involvement and it’s difficult for us to be involved with those projects on a longer term.

CCG M, interview 5

The working relationship between external BI staff and the CCG could negatively influence the way the BI unit was able to affect CCG critical review capacity, and prevented it from being fully integrated into decisions about evaluating services:

I probably would say it’s this lack of trust and respect from the CCGs . . . It just feels like actually no one respects our opinion when we give it to them. You know, ‘you asked us to do a piece of work. We explained to you in detail why it’s not a good piece of work to do and why it’s not useful and how it’ll actually probably muddy the waters more than it will help’, and then you still end up doing it. It makes the working relationship between CCGs and the CSU quite difficult I think.

CCG M, interview 2

Problematically, this tension could be related to the system pressures faced by CCGs, which are not focused on long-term evaluation of projects because of immediate demands for service provision and cost reductions:

I think there is a lack of strategic planning at the CCGs. I think we show the commissioning cycle in terms of being able to evaluate need and you do something about that, you commission services differently and then you evaluate that, but I’m not sure that that happens . . . What they’re bothered about is actually implementing these schemes and doing that commissioning, but what they don’t think is ‘how am I actually going to evaluate whether that’s working or not?’.

CCG M, interview 1

In conclusion, the external BI teams, acting in a consultancy capacity, were able to provide a co-ordination capability by drawing together multiple, diverse information sources to provide strategic insight to CCGs, beyond standardised data reporting. However, at times there was a lack of engagement between the CCGs and the external BI teams. This could undermine critical review capacity, as a result of the influence of socialisation capabilities. On many occasions, exploitation processes within CCGs were seen as underdeveloped, and evaluation of services was a secondary consideration in many cases.

Conclusion

Business intelligence units, in general, have the potential to enhance the critical review capacity of CCGs by acting as a co-ordination capability. However, they operate at two distinct levels: embedded teams and external teams. Although embedded teams were able to overcome socialisation limitations by interacting with commissioners on a relational basis and using local knowledge to enhance assimilation and transformation of knowledge, they reported restrictions of systems capabilities. Conversely, external BI teams were better equipped to overcome systems limitations, but were influenced by socialisation barriers as they were perceived as being ‘outside’ the CCG. Exploitation processes of BI teams were underdeveloped at both levels.

Sharing strategy: bridging health and social care

Integrated teams, particularly those that straddled health and social care, were the focus of four of the CCG cases, but were represented in some form across all 13 CCG networks. Spanning two or more organisations, integrated roles were designed to acquire and integrate information from disparate information systems, overcoming systems capabilities. In particular, barriers between health and social care information systems were identified as limitations of knowledge acquisition and sharing among CCGs, and integrated roles were perceived as a way to overcome these barriers:

It has specifically improved communication a lot between social services and health I think. Because we both have two different computer systems we don’t talk to each.

CCG H, interview 1

This integration was seen as key for commissioners attempting to design services, as integrating perspectives and demands from multiple organisations could be seen as problematic:

It’s challenging to get that shared interpretation of what the information actually means for actual service design because clearly there are different interpretations you can apply to the same information. From a commissioner perspective, we will see a problem or challenge from one particular perspective, but health-care providers will see a very different challenge, and so we will support different service interventions.

CCG D, interview 12

By facilitating the acquisition of diverse sources of information, integrated roles enhanced the assimilation and transformation of multiple forms of knowledge, enhancing critical review capacity of the CCG. This could enhance the commissioning process and have a positive influence on patient care, acting as a co-ordination capability. This was particularly the case for complex care provision for older people, who were often admitted to hospital in the absence of effective collaboration across health and social care organisations. Integrated care teams encouraged knowledge mobilisation across sector boundaries, enhancing service provision:

It [an integrated health and social care pathway] represents a smoother pathway for the patient. In the past the older patient would have been taken into hospital, the patient discharged, and community social care teams have little contact with what’s going on. The older patient may then be subsequently re-admitted to hospital because the social care support wasn’t there. Because we’re integrated now we can see two sides, the need for social care, as well as health care, for the patient, and so prevent readmission.

CCG F, interview 5

In particular, being able to take a wide view of the patient case enabled integrated teams to move past some of the potential challenges of systems capabilities, such as distinct funding streams for social and health care, which complicates elderly care provision:

I think obviously the funding bit, social care funding and health funding. They’ve both got their pots and they both want to protect their money and ‘no, that’s not my job, that’s health.’ ‘No, that’s not us, that’s social care.’. I’ve really just been, you know, in the middle saying ‘the person needs care. Let’s put it in and fight about it later . . . Let’s just get this care in. It’s Friday afternoon. Let’s argue about it next week’, which is how I think it’s got to work. There’s got to be somebody to draw the line, hasn’t there? I know there’s cuts and problems with funding, but the patients have got to have the care, haven’t they?

CCG H, interview 9

In one of our cases, the CCG had almost fully integrated its resources with those of the local authority. Working in the same location, sharing staff in hybrid roles (those who were employed 50% by the CCG and 50% by the local authority) and combining budgets for health and social care were seen as key mechanisms through which the CCG provided a more joined-up approach to service design:

I think I’m very, very pleased to have found that [the] CCG is able to work so closely with [the] borough council to the extent that we’re actually based in council buildings as you can see. That was never the case in the days of PCTs. There was a lot of suspicion between the two, but now they work hand in hand. That I think facilitates what you’re talking about – being able to bring in social care so that we can give people a package of care that’s not fragmented.

CCG F, interview 10

Integrating teams created a more relational approach to joined-up commissioning between commissioners and the local authority, rather than a transactional approach in which funding was a major barrier. In addition, an integrated approach positively affected the alignment of other stakeholders. For example, as a result of close working relationships between the CCG and the local authority, commissioners reported that the secondary care organisation was more ‘in line’ with the shared vision of the commissioning network:

To actually make all the changes that we’re going to make it’s required a lot of engagement between clinicians and directors of both CCG and the borough council and secondary care and that’s only happened because the CCG and the borough council work very closely together and there’s no suspicion anymore, so everyone’s pushing in the same direction and because of that I think that secondary care feel that . . . so you can actually get people in a room that actually will agree to things more readily.

CCG F, interview 10

The integrated teams acted as a co-ordination capability for the CCG network by encouraging the sharing of information and resources across boundaries, mediating the barriers of systems and socialisation capabilities. Although this facilitated commissioning decisions, the more important impact of integration was seen in the transformation of service decisions, having a positive impact on complex patient care:

I had a 100-year-old lady this week . . . I think third case of pneumonia, doctors wanted to admit her. She didn’t want to be admitted. She wanted to stay at home. In anybody’s eyes that’s a health need because there was a hospital admission . . . The matron phoned me up and said ‘what can we do? What can you do? Do you think we ought to put care at home in?’ but when I looked at the situation . . . I could have gone to the care at home team and put a team in to support her at home, but actually she was having agency care. Really good package, carers she trusted. They all really cared for her. You know, they were first name. Really good and it worked really well. So instead of taking them out and ‘it’s a health need. Let’s put health in to do it’, I left them in, looked at the contingency [fund], we had some money in the contingency, increased the care visits so they could spend a lot of time with her. So perhaps if that was cut and dried I should have used health because that was a health need, but I used social services because that was for her . . . So I think there’s ways to work round things.

CCG J, interview 9

Boundary spanning: engaging with the voluntary sector

One CCG case we explored had an integrated team that worked to mediate systems barriers between health and social care, but also encouraged integration from external providers, most notably the voluntary sector. In particular, the integrated team used voluntary services to enhance acquisition of information about complex patients, influencing transformation into better services for patients:

We’ve been dealing with cases over the last 2 or 3 days in fact where volunteers obviously are our eyes and ears into some of these people as well, which is very useful and also because they go in once or maybe twice a week and they’ll sometimes see things that even if that person’s having a daily care package people don’t always see. So the volunteers can feed that back to the [integrated team] who will have a good helicopter view of the services they need.

CCG J, interview 2

The integrated teams gave a number of examples of when they had engaged with voluntary services to provide patients with services not covered by health or social care teams, or to extend the services available. The following illustrates the importance of these services:

One of the nurses, she’d been out to a gentleman . . . He’s got a leg wound and it’s got maggots in it. So she’d come back to the office to me and said ‘oh, can we get some care in there because this gentleman’s really in a state? There’s flies everywhere in the house, it’s a tip. You know, he’s not managing, he’s not coping. Can we get some care?’. I said ‘OK. . . . He’s back on his feet, care’s ended, nursing needs have ended, he’s fine’ . . . So that was a joint effort by all – social services, the nursing team and the voluntary service. All three involved there.

CCG J, interview 10

In the example above, the integrated team acted as a co-ordination capability, bringing together health services, social services and the voluntary sector. However, by engaging with the voluntary sector at an early stage, the integrated team limited the number of services required by the patient. The volunteer was able to engage with the family to improve the patient’s standard of living, resulting in a better health outcome, removing the need for a social services care package and potentially avoiding a crisis situation that could have resulted in an urgent care admission. Other examples of volunteer involvement also noted the importance of creating a safer living environment for preventing accidents, ill health or urgent admissions:

I mean just decluttering somebody’s house – the impact that has on unplanned hospital admission is immense because if somebody doesn’t fall over they don’t end up sat waiting for the ambulance service to come and get them for 7 hours, they don’t end up queuing up at A&E [accident and emergency], they don’t end up in hospital, they then don’t end up stuck there for 3 weeks because social care can’t pick up what their needs are. Just through simply having somebody spending an hour chatting to them about making their house a bit safer.

CCG J, interview 2

A further benefit of engaging with the voluntary sector prior to admissions was an awareness of the potential of voluntary services in facilitating discharge from hospital after an admission:

If somebody is admitted in an unplanned way into hospital it means that the social care system for that person at home has to change in order for them to come out of hospital and the complexities around doing that are just immense. The other side to it is also about trying to understand what we can do to better support people that are making a planned admission to hospital so that discharges aren’t delayed, so that people make a sustainable return home . . . What we can do is just make sure we’ve got a volunteer going in working with the patient to prepare their home – so sort out any trip hazards, make sure they’ve got kind of heat and light available when they come home, make sure they’ve got basic groceries, have got contact lists, all those sorts of things – and that volunteer will then work with the patient to get them into hospital and work with the patient when they come home from hospital so that that whole pathway is delivered more effectively.

CCG J, interview 3

In other examples, voluntary services were used for more-relational interventions, such as befriending, to prevent social isolation. In many cases, the integrated team worked as a co-ordination capability to facilitate the involvement of volunteers and to integrate them with the health or social care services also being provided, to develop a more holistic package of care:

One was for a befriender for a lady who’s housebound, had recently been bereaved and was very, very withdrawn and desperately needed friends. No family, no children, her husband had obviously died and she was really quite sad. The nurses couldn’t do anything to help really. They’d done their piece as far as medical, so they sort of said to me and at the same time it was discussed at the risk profiling as well for this lady what can we do, so I said ‘well, I can ask [voluntary organisation] if they have a befriender’.

CCG J, interview 5

The above examples illustrate instances in which voluntary services have been used as an extension of, or have been integrated with, statutory health and social care services. Facilitated by the integrated teams, which co-ordinate the interaction of all three organisations, involving volunteers in this way has the potential to improve quality of life for patients. In some circumstances, this can be identified through improved health outcomes or a reduction in health and social care services, or through more-relational interventions such as befriending or social activities.

A word of warning: ‘patching’ health and social care

Although integrated teams were a key co-ordination capability for CCGs in encouraging the integration of knowledge, but also in ensuring comprehensive packages of care for older people, we also uncovered examples of situations in which voluntary services had been used to ‘patch’ gaps in health and social care. In some cases, illustrated in the following quote, this was because of limitations of funding or responsibility for the patient:

Not long after I started I had a case with a lady . . . I think she was due to go in for some surgery and she needed an MRSA [meticillin-resistant Staphylococcus aureus] wash each day and she could do a certain percentage of her body but she couldn’t do her back. So that was basically all it was and health-care were saying ‘it’s not us, it’s not a health need, it’s a social care need.’. Social care were saying ‘no, it’s not us, it’s a health need’, and this went on for a couple of hours and then in the end I just sort of went back to her and I said ‘Look, we’re having a bit of difficulty getting somebody in to you. Would you be prepared to pay, you know, it could be £5 each day for the next week for somebody to come in from a voluntary organisation and support you with this?’.

CCG J, interview 4

Although voluntary services could be effective in preventing patients from falling through the ‘gaps’ in service provision, there were also examples of situations in which they were used because of lack of capacity within the health and social care systems. Owing to pressures on, particularly, social care, there were examples of voluntary services being used instead of statutory services:

So we use them mainly because social services are so stacked up accepting our referrals that we’ve used the [voluntary organisation] more because they’re an agency that can . . . They can do kind of everything in one visit. They can spend up to 2 hours with somebody. I mean this is something which in the past social care were able to do . . . but now as everything is moving so quickly and everyone’s got so many patients there’s not enough time.

CCG J, interview 5

We do not suggest that voluntary organisations should not be ‘patching’ health and social care services, as they can play an important role in crossing organisational jurisdictions to stop patients falling through gaps. However, there were examples of voluntary services being used to fulfil gaps in services as a result of a lack of capacity within health and social care teams. The use of voluntary services in this way can be beneficial to patients, preventing crisis admissions and enabling them to remain at home, but it should be negotiated carefully.

Conclusion

Integrated teams acted as a co-ordination capability in two ways. First, they encouraged the integration of knowledge from two (previously disparate) data systems. By integrating health and social care information, hybrid roles could develop transactional relationships to overcome systems limitations and positively influence the commissioning process and patient services. Second, by taking an overview of the wider context, integrated teams developed relational roles that encouraged the acquisition of external information, such as from the voluntary sector. This could then be used to further enhance the commissioning process and facilitate the delivery of comprehensive care to complex elderly patients.

Introduction

Based on the qualitative study and extensive literature research, we designed a self-development tool for CCGs and their partner organisations. The tool has been constructed in line with recommendations for a high-standard psychometric tool. 124,125 It has been cocreated with members of the PPI and pilot tested with a volunteering CCG to enhance face validity and applicability. The next step is to provide evidence for construct and criterion validity, which involves collecting data from CCGs and testing the psychometric properties of the tool.

The self-development tool captures ACAP, which is the mechanism linking combinative capabilities of ACAP (antecedents to ACAP, including system capability, co-ordination capability and socialisation capability) and its outcomes for the CCGs. The tool focuses on how different parties in a commissioning process acquire and use knowledge, and how they collaborate within CCGs and across partners to make decisions.

The aim of the tool is to provide CCGs with information on their commissioning processes, which can help to improve their decision-making. The self-development tool offers the following benefits: first, CCGs interested in improving their decision-making processes can do so with the help of a structured tool, which provides specific feedback on a number of vital aspects of the decision-making process. The feedback provides an evidence base for participating CCGs, facilitating discussion and communication with different stakeholders in a targeted way. Second, over time, data collected with the tool form an anonymous database. This will provide statistical norms and benchmarks with which CCGs can compare their scores, so that they can identify their specific strengths and potential areas for improvement. Third, CCGs will be able to identify how the commissioning process is perceived from the perspectives of different stakeholders.

The self-development tool comprises a 15-minute online survey, available to all CCGs and their partner organisations free of charge. A participating CCG will invite all individuals working directly for it, as well as its partners. The more individuals who complete the tool, the more detailed the feedback will be, providing a more valid evidence base for the CCG to work with. An anonymised report is generated from the responses, which includes a summary of the main constructs and feedback on the average scores, alongside recommendations on how to improve (see An example of a Clinical Commissioning Group report). Over time, with increasing CCGs participating, the database built will provide norms and standards, which provides a basis for comparison.

The RQs guiding the tool development are the same as those for the qualitative study. Our overarching RQ is as follows: how can we enhance ACAP of CCG-led commissioning networks in health care to inform decisions to reduce potentially avoidable elderly care acute hospital admissions? This breaks down into three linked RQs:

1. How do CCG external relationships and system antecedents to new commissioning arrangements affect CCG critical review capacity?

2. What features of CCG organisational context (i.e. internal capacity) affect CCG critical review capacity?

3. How do we reduce the variation between PACAP and RACAP?

Theoretical justification and brief overview of constructs

This section discusses the constructs in the tool and provides the rationales to include them. As shown in Figure 2, the tool consists of the four elements of PACAP and RACAP (acquisition, assimilation, transformation and exploitation), linking relevant antecedents (organised into three categories: socialisation capability, system capability and co-ordination capability) with a number of commissioning outcomes of ACAP following QIPP (including commissioning-related criteria such as the perceived quality of the commissioning, as well as organisational outcomes such as innovation, productivity and prevention). Most of the variables are assessed with respect to one particular exemplary commissioning process, which is hereafter referred to as the commissioning of Project X, whereas the remaining variables are about the CCG in general.

FIGURE 2.

Overview of the framework.

In the following section we provide a brief explanation for each group of variables. An extensive description of the development of all variables, including their theoretical and empirical origins, can be found in Appendix 1. Table 4 summarises the variables and items of the tool.

Methods

The development of the tool started after the qualitative part of the study finished and all executive reports were submitted. We developed the tool in three iterative steps in collaboration with PPI. The fourth step of the tool development, the rigorous psychometric testing of it, lies outside the scope of this bid and is currently under way.

1. Development of the theoretically guided model, including antecedents, ACAP and consequences, based on previous theory and research as well as the results of the qualitative study (in close collaboration with PPI).

2. Operationalisation of all three elements of the model (antecedents, ACAP and consequences). We developed items following our theoretical assumptions and by drawing heavily on existing measures, amending and extending those based on the results from the qualitative data.

3. Revising and editing items based on feedback from members of PPI, members of a CCG, academic experts and conferences. The tool was refined with comments from the PPI group and conferences attended, before sending out for piloting in one participating CCG. The role and involvement of the PPI is described in depth in Chapter 10.

4. Psychometric testing (outside the scope of this project): we are recruiting CCGs for participation to collect sufficient data to rigorously test the psychometric properties of the tool and to amend where appropriate.

The final version of the self-development tool consists of an online survey, which takes ≈15 minutes to complete. The survey should be self-explanatory and the first page explains what the survey is about and how to complete it. It also asks for respondents’ consent, guaranteeing confidentiality. Ethics approval to pilot the tool and to gather data in the future was sought and granted by the University of Warwick and the Biomedical & Scientific Research Ethics Committee.

During the period of development, we presented the tool in several conferences for feedback and to advocate its future dissemination. The conferences and events included Commissioning LIVE London (2 March 2017, London), Improving the Capabilities of NHS Organisations to Use Evidence Workshop (2 May 2017, Warwick) and Health Service Research UK (6/7 July 2017, Nottingham). As a result, we were in conversation with a number of CCGs about the tool and advocated the publicity of the tool. We have also been in conversation with NHS England about the potential dissemination of the tool.

We have also been in contact with CCGs for pilot sites, raising the awareness of the tool. All CCGs that we approached were sent an information sheet (see Appendix 2) and consent form (see Appendix 3), introducing the tool and guaranteeing the confidentiality of participants. Participants and CCG anonymity will be maintained at all times during the dissemination of the research findings.

An example of a Clinical Commissioning Group report

This section shows results from our pilot CCG and its partner organisations as an example of the nature of evidence that CCGs can receive from the tool. It is important to note that the psychometric testing of the tool is still outstanding, hence this chapter exemplifies what the tool will be able to demonstrate (1) once a database has been built and (2) when a representative number of CCGs participate. The data analysis again follows the conceptual framework depicted in Figure 2 and provides means and standard deviations (SDs) of combinative capabilities, ACAP and outcome of commissioning. The mean is the central or typical value in a particular set of data, whereas the SD shows how much respondents differ from the mean value of the group, that is, the average score of respondents in this group. A mean score of three signifies the scale’s conceptual midpoint. A low SD suggests that respondents’ views are close to the mean of the data set, whereas a high SD indicates that respondents’ views are scattered. Once the data set has been built, we will be able to integrate and compare CCGs’ scores with the population norms.

We present the results of each category in radar charts for easier comparison, and means and SDs are summarised in Table 5. Owing to the small sample size of our pilot study, we are unable to provide significance testing to demonstrate differences between groups or variables. Hence, when describing a variable as higher, lower, less likely, etc., this does not indicate statistical significance and is merely descriptive at this point.

Table 5 is exemplary of what a report back to the CCG might consist of; however, the very low number of participants in this pilot group does not allow for indicative assumptions.

Socialisation capability

As shown in Figure 3, we grouped all constructs from socialisation capability except trust and emotion in one radar chart for easy comparison. The results show that CCG staff perceive themselves to strongly identify with, and to be committed to, the CCG, sharing knowledge among themselves as individuals. They also report feeling heard and voicing issues, but may feel less comfortable with taking charge during the commissioning process. Nevertheless, a high SD of the construct ‘taking charge’ shows great variation in individuals’ responses (see Table 5 and Figure 3).

FIGURE 3.

Socialisation capability (CCG staff, mean): range from 1 (strongly disagree) to 5 (strongly agree). KS, knowledge sharing.

We also assessed some constructs of socialisation capability that are related to the commissioning work with the CCG (see Table 5 and Figure 4). Both the mean and SD of the construct ‘being heard’ are lower than for other constructs, suggesting that partners believe that the CCG disengages with comments and feedback from partners.

FIGURE 4.

Socialisation capability (partner, mean): range from 1 (strongly disagree) to 5 (strongly agree). KS, knowledge sharing.

The results show that ‘trust in partner that should have worked with more closely’ is relatively low compared with ‘trust in partner most crucial to outcome’, ‘trust in CCG managers’ and ‘trust in CCG colleagues’ (Figure 5); however, a high SD of this variable (see Table 5) shows that respondents’ views on this matter varied quite significantly. Moreover, a low SD (see Table 5) of ‘trust in CCG managers’ suggests that most CCG staff have quite a high level of trust in CCG managers.

FIGURE 5.

Trust (CCG staff, mean): range from 1 (strongly disagree) to 5 (strongly agree).

We also assessed 12 emotions related to the decision-making process of both CCG staff and partners in the survey (six positive and six negative emotions). Emotions are a good indicator of the decision-making process. As presented in Figure 6 and Table 5, the data show that CCG staff mostly show positive emotions and occasionally feel angry and surprised during the commissioning process. Interestingly, SD of emotion shows that CCG staff vary significantly in the frequency of negative emotions they experience.

FIGURE 6.

Emotion (CCG staff, mean): range from 1 (never) to 5 (very frequently).

Our data show that positive emotions experienced by partners are in a similar pattern to CCG staff; however, partners experience fewer negative emotions than CCG staff (Figure 7). Nevertheless, the high SD of negative emotions such as ‘angry’, ‘frustrated’ and ‘fed up’ suggests that some partners experienced these three emotions frequently or very frequently during the commissioning process (see Table 5).

FIGURE 7.

Emotion (partner, mean): range from 1 (never) to 5 (very frequently).

System capability

System capability includes the following three constructs: knowledge-sharing platform, feedback seeking and organisation/team infrastructure. As shown in Figure 8 and Table 5, the relative high mean and low SD of the construct ‘organisation/team infrastructure’ suggests that most respondents believe that there is adequate and sound organisation/team infrastructure in place. Meanwhile, the low mean and high SD of the variable ‘feedback seeking’ show that respondents vary significantly in their views of how frequently CCGs seek feedback from partners and customers.

FIGURE 8.

System capability (CCG staff, mean): range from 1 (never) to 5 (very frequently). KS, knowledge sharing.

Co-ordination capability

There are two constructs, cross-functional interaction and participative leadership, in co-ordination capability. Figure 9 and Table 5 draw the mean and SD of these two constructs. The high mean and low SD of both constructs suggest that most respondents are quite confident that (1) the CCG general manager shows the characteristics of participative leadership and (2) there are cross-functional interactions within the CCG.

FIGURE 9.

Co-ordination capability (CCG staff, mean): range from 1 (never) to 5 (very frequently).

Absorptive capacity

The data in Figure 10 show that the pilot CCG frequently engaged in acquisition and transformation, with slightly lower frequency in assimilation and exploitation. Moreover, low SD across all four elements of ACAP shows that responses from CCG staff are close to the mean values, that is, they are frequently engaged in the four elements of ACAP (see Table 5).

FIGURE 10.

Absorptive capacity (CCG staff, mean): range from 1 (never) to 5 (very frequently).

Outcome of commissioning

Outcome of commissioning includes the following four constructs: CCG innovativeness, quality of decision, use of resources and acceptability of outcome. As shown in Figure 11, innovativeness, quality of decision and use of resources were rated highly. Yet most respondents rated the acceptability of outcome as low, and its high SD shows great variance in responses in rating this construct (see Table 5). This result raises the question of why the acceptability of the outcome is rated so low. As a consequence of this result, we have added an open comment box to the tool, asking participants to clarify.

FIGURE 11.

Outcome of commissioning (CCG staff, mean): range from 1 (poor) to 5 (excellent).

Similar results are shown in the partner survey, that is, most respondents rated the acceptability of outcome as low (Figure 12); however, the relative low SD of acceptability of outcome from partners suggest that partners’ answers to this question are quite close to each other, that is, low acceptability of outcome (see Table 5).

FIGURE 12.

Outcome of commissioning (partner, mean): range from 1 (poor) to 5 (excellent).

To sum up, this section analyses responses from both partners and CCG staff on the following aspects: socialisation capability (with a special focus on trust and emotion), co-ordination capability, system capability, the four elements of ACAP and outcome of commissioning. Nevertheless, the analysis should be interpreted in light of the limited number of respondents in this pilot group (10 partners and 8 CCG staff), as the high SD of certain constructs might be a result of the limited number of informants taking part in the survey.

Dissemination of the tool

We plan to disseminate our research in the following ways. First, we have obtained an Economic and Social Research Council acceleration grant to disseminate our research and enable evidence use in health-care commissioning together with another NIHR-funded research project at Warwick Business School. The objective of the grant is to convert academic findings of the two research projects into practice and make the evidence more accessible to practitioners. In doing so, we build an online professional development platform, which offers integrated resources for better commissioning decision-making. Specifically, the online development platform includes, but is not limited to, the self-development tool, case studies of the two studies, discussion boards, a process roadmap and links to relevant resources and blogs.

Moreover, NIHR Collaborations in Leadership for Applied Health Research and Care (CLAHRC) West Midlands (WM) and East Midlands (EM) (a translational health research initiative funded by NIHR) could be utilised to disseminate the tool, given the close relationship with NIHR CLAHRC WM and EM via steering group members. A possible way to engage might be to raise awareness of the tool in their newsletters and invite friendly CCGs from the WM and EM to participate in the self-development tool and write testimonials about the usefulness of it.

Introduction

This chapter reports on the PPI undertaken in this study within work package 3, focuses on the discussions of the PPI reference group and explores the nature of their contribution to the study.

The importance of PPI in research is briefly introduced in the context of the wider PPI evidence base. The reporting of PPI within this chapter is guided by the Guidance for Reporting Involvement of Patients and Public (GRIPP) 2 Short Form checklist to ensure that important aspects of PPI are considered. 145 Thus, the chapter includes aims, methods, results, discussion in relation to outcomes and discussion in relation to critical reflection from the study of what went well and what went less well.

Methods

Reference group discussions

Reflection on process

We have included two researcher reflections and two public member reflections on the PPI in the study.

Discussion

The PPI in this study aimed to provide the critical review capacity in relation to study results, specifically to discuss emerging findings. The reference group provided an effective context in which to discuss how commissioning was operating, to create new challenges and to discuss the validity of identified barriers. As such, they provided a form of community validity. The study also aimed to develop understanding of the potential for PPI to contribute to organisational studies in health care. The contribution and impact clearly demonstrate the significant potential for PPI to grow in such studies. A key area of contribution was the development of the ACAP tool, particularly in relation to the dimensions and items, in addition to the practical utility. The research team will be working with the reference group to publish the PPI aspects of the study.

Implications for practice

We have noted that our study of CCG-led commissioning networks has taken place soon after their inception. In part, this might explain why their critical review capacity for knowledge mobilisation (in our conceptual terms, ACAP) is not well developed. Over time, we might expect relationships with key stakeholders, such as GPs, PPI representatives and CSUs, to further develop to the benefit of critical review capacity for knowledge mobilisation. During the study period, however, CCGs exhibited less concern about their lack of critical review capacity than we might have expected (as evident in challenges we experienced in accessing CCGs described in our research design). To encourage CCG-led commissioning networks to take their critical review capacity more seriously, as a first practical implication, we offer our psychometric ACAP tool in the hope that this encourages CCG-led commissioning networks to develop their critical review capacity, not just regarding our tracer study domain but more generally across all service domains. The tool is orientated towards bridging the gap between their PACAP and RACAP, so that the knowledge acquired and assimilated by commissioning networks is transformed and exploited, and consequently evidence-based service interventions are scaled up.

In our original protocol, our fourth RQ was ‘How can we ensure that PACAP is realised in commissioning networks?’. We suggest that our original use of the word ‘ensure’ was overoptimistic; the study shows how multiple factors influence the exploitation of new knowledge, and makes a range of suggestions about how the likelihood of transformation and exploitation could be increased, as described in the following paragraphs.

Policy-makers’ influence on the ACAP of CCG-led commissioning networks most obviously lies with influence on their systems capabilities. Policy-makers mandate the performance and incentive structures to which CCG-led commissioning networks respond. In some respects, policy-makers can be seen to support ACAP through, for example, requirements that CCG-led commissioning networks encompass PPI. However, policy requirement is one thing, its implementation is another. PPI was patchy across the CCG-led commissioning networks that participated in our study. Similarly, GP involvement was not as it might be. The relative absence of both across our CCG-led commissioning networks had deleterious effects on ACAP. CCG-led commissioning networks appeared to be more orientated towards what might be regarded as transactional mode with providers regarding contracting. They were thus concerned with monitoring contracts rather than with behaving in a more strategic way towards use of evidence in re-engineering services. Any vision of ‘world-class commissioning’ from policy-makers appears to be slow to realise. Policy-makers might attend to such matters.

In part, the above might link to socialisation capabilities, specifically the managerial cadre employed by CCGs, which appears wedded to its traditional bureaucratic orientation. We caution that this is merely a tentative suggestion derived from impressions during fieldwork, which might be followed up in further research. Nevertheless, to realise policy aspirations for evidence-based commissioning might require greater sophistication on the part of managers. In particular, managers might reflect on their involvement of GPs, who may be decoupled from their practice-based colleagues, and thus need to consider how commissioning-based GPs and their practice-based peers might be better linked to engender co-ordination capability that enhances ACAP.

Alongside this, CCG-led commissioning networks themselves might enhance their ACAP, not only through co-ordination capabilities, such as more effective PPI and GP involvement in their commissioning, but through better use of and interaction with those providing BI, either internal or external (CSUs) to the CCG. Admittedly, our study was limited, but it is striking that ACAP was enhanced when the BI function was ‘in-house’. In this case, interactions between data analysts and commissioners were relational rather than transactional, the former leading to enhanced service from the BI function to commissioners, as they understood the data needs of commissioners in real time. The challenge for CCG-led commissioning networks that rely on external CSUs is to replicate this relationship.

Better relationships across the network of service providers and commissioners are certainly a significant factor in enhancing ACAP. Often, the problem is that different perspectives need to be bridged, for example between health and social care, so that they collaborate more effectively. Voluntary sector organisations help provide the necessary social integration to support this. More generally, shared strategies and mental models of care across organisations, managers and professionals from different providers and commissioners enhance the ACAP of the CCG-led commissioning network.

Transferability of analysis

Although we highlight specific implications for elderly care above, we note that this is a tracer study designed to surface issues, lessons and a self-development tool for allowing commissioning networks to use different evidence across all service domains. In these terms, our study indirectly affects improved service and potential reconfiguration of pathways; that is, our intention is to directly affect the use of evidence rather than the service outcome itself. Nevertheless, our study, which provides insight into the application of ACAP, is particularly relevant to current health-care settings. Health-care settings globally are subject to financial parsimony, and, as such, need to be smarter about knowledge mobilisation, in particular how acquired knowledge informs the planning of, and funding for, service interventions, not least given burgeoning demands as a result of the increasing demands from older patients on hospital services (e.g. in NHS England, ≈50% of hospital beds may be occupied by older patients). For example, a health-care organisation may invest significant resources in IT capacity to acquire information, but if this information is not then used intelligently to inform service development, investments will merely result in PACAP, rather than RACAP,13 without patient, financial or competitive advantage. To emphasise, we note that our analysis is not limited to commissioning the care of older people, as application of the ACAP concept allows theoretical generalisation to other domains of health care, both the delivery and commissioning of services, and indeed other public services characterised by professional organisation and central government intervention. 108

Further research

Finally, in terms of further research, co-ordination capabilities within health-care organisations have been identified within the literature, as represented by the following: the development of learning relationships through the establishment of internal and external networks, staff development and training, appropriate leadership, organisational strategy, investment in information support systems, participation in decision-making and, more generally, social relations inside and outside the organisation. 33,34 Although our empirical study has highlighted professional involvement, client involvement and integrated service delivery models, others may wish to examine possibilities offered by other forms of co-ordination capability.

We have noted that our study took place at the early stages of development of CCG-led commissioning networks. We suggested that relationships to allow GP involvement and PPI to flourish would take time to develop, although we noted that front-line GPs seemed to be disengaging with the new commissioning structures and processes. In short, there is a temporal dimension that is relatively absent in our analysis, but that further study might examine. As CCGs developed, were some of our concerns addressed, particularly those about GP involvement and PPI?

Contributions of authors

Graeme Currie led the research and writing of the report.

Charlotte Croft and Yaru Chen carried out fieldwork and cowrote the report.

Tina Kiefer led the development of the psychometric tool derived from fieldwork.

Sophie Staniszewska led development for the governance of PPI.

Richard J Lilford provided methodological advice for fieldwork and the psychometric tool, contributed to the analysis of empirical findings and checked and edited the final report.

Publication

Croft C, Currie G, Staniszewska S. Moving from rational to normative ideologies of control over public involvement: a case of continued managerial dominance. Soc Sci Med 2016;162:124–32.

Data sharing statement

Owing to the nature of this study, the type of data collected and the conditions attached to ethics approval, there are no data available for wider use. The empirical sections of the report contain a rich section of primary data quotations across all 12 comparative cases. All queries should be submitted to the corresponding author in the first instance.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HS&DR programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HS&DR programme or the Department of Health and Social Care.

Socialisation capability

System capability

Co-ordination capability

Absorptive capacity

Outcomes of absorptive capacity