NHS Commissioning Practice and Health System Governance

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 08/1806/262. The contractual start date was in December 2009. The final report began editorial review in June 2013 and was accepted for publication in February 2014. 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

none

Copyright statement

© Queen’s Printer and Controller of HMSO 2015. This work was produced by Sheaff et al. under the terms of a commissioning contract issued by the Secretary of State for Health. 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.

Background

Few questions are more important to the NHS than how its commissioners exercise governance over local health economies. Commissioners pay for health care on behalf of patients who cannot do so themselves and on the state’s behalf; in the absence of public ownership and direct managerial control by the state, they also exercise governance over the service providers. This study aims to examine the means, contexts and effects of commissioning practice that was current in 2010–12, that is the activities of assessing health needs, selecting and contracting providers to meet them, monitoring the outcomes and then repeating the cycle. 1,2 We focus on the ways in which health-care commissioners can influence health-care providers within a quasi-market health system.

NHS commissioning: the policy context

Policy initiatives related to commissioning since 1991 have set the basis for current NHS commissioning policy and the system introduced in April 2013. Since 1991, three distinct main commissioning structures have evolved:

1. Population-based commissioning. A single body commissions health services for the entire resident population within its geographical boundaries. This structure includes public health activity, for evaluating population health-care needs and initiating preventative activities. In England, District Health Authorities began commissioning services for populations of 200,000–500,000 people in 1991.

2. General practice-based commissioning. General practices, or another gatekeeper and budget holder, individually or collectively commission services, the general practitioners (GPs) acting as proxies or advocates for their registered patients when making referral decisions. These commissioners tend to serve perhaps 5000–100,000 patients. GP fundholding was the best-known English variant.

3. Client-based commissioning. Patients themselves choose a health provider, which a commissioning organisation then pays on their behalf. Consequently the provider has to be paid per episode of care; in most health systems, through a diagnosis-related group (DRG)-based tariff system. Another variant is to give patients a voucher or budget to pay for care. Client-based commissioning is still an emergent structure in England, represented by the Patient Choice policy, personal health budgets and the ‘payment by results’ (PbR) system (see section Phase 5: 2006–10 – client-based commissioning).

Clinical Commissioning Groups (CCGs), introduced in April 2013, are essentially a variant of population-based commissioning, but also resemble general practice-based commissioning in that GPs play a pivotal role in their governance. One can distinguish six phases of the evolution of NHS commissioning in England.

Media of power

The ‘therapeutic state’ (p. 254)40 co-opts and adapts the ‘technologies of power’ that it believes will reinforce its control over the population, promote and implement policies (e.g. regarding population health, reproduction, the control of deviancy) and discipline the medical profession accordingly. Foucault argued41 that control within and between organisations occurs through a dispositif: a structured complex of diverse, coexisting technologies of power42 including professional disciplines, surveillance, task sequencing, task distribution, coercion and panoptical control. 49,50 Commissioners generally try to exercise governance over health-care providers by combining several methods in parallel. 51 We call each such method a ‘medium of power’, because each embodies a collection of ‘technologies of power’. 52 These complexes of technologies of power are historical ‘positivities’ that can be identified only empirically, case by concrete case. 53 International overviews35–37 suggest that health-care commissioners generally use one or more of six media of power: managerial performance of commissioning; negotiated order; discursive control; resource dependency and financial incentives; provider competition; juridical control.

Modes of commissioning

As noted, commissioners try to exercise governance over providers through particular combinations of media of power. We call each such combination a ‘mode of commissioning’. Globally, many different modes of commissioning are found. The above research findings suggest, in sum, that differences in modes of commissioning help explain:35

1. patterns of provider development, the spread or absence of specific kinds of provider or services; corporatisation and concentration of capital

2. health systems’ capacity for cost control

3. the development and use of evidence-based medicine

4. patterns of managerial development of commissioning and medical involvement therein.

Different types of commissioning organisation are likely to develop different modes of commissioning, whose effects partly depend on how providers react. 24 Except in the extreme case of ‘gridlocked’ governance,189 each mode of commissioning leaves providers some room for manoeuvre. Each medium of power might therefore be expected to have different effects even on similar providers (e.g. university hospitals) according to the institutional context. One might expect different modes of commissioning to develop for, say, diverse care groups. This brings us to our research questions.

Research design

Our research design was a multiple mixed-methods191 realistic evaluation. 192 Its components were:

1. Content and discourse analysis of policy documents and interviews with policy-makers and managers to elicit their programme theories of NHS commissioning, answering RQ1 by identifying their understanding of:

   1. the intended policy and service outcomes of NHS commissioning

   2. the mechanisms that would produce these outcomes.

2. A cross-sectional analysis of publicly available managerial data about local health economies (commissioners, providers, socioeconomic context), testing for any associations between commissioners’ characteristics and the policy outcomes identified in the programme theory of NHS commissioning. In agreement with the Health Services and Delivery Research (HSDR) programme, this analysis was regarded as an initial exploration of the value and uses of reanalysing published managerial data to characterise and evaluate the impacts of NHS commissioning. The unit of analysis (‘local health economy’) would for practical purposes be the PCT, considering what data would probably be available. This method contributed to answering RQ3 and RQ5.

3. A systematic comparison of case studies of commissioning in five English case study sites, with induction of common patterns, and exploration and explanation of contrasts. Within each site these case studies were longitudinal, tracing the formation of commissioning structures and practices in recent years. The framework structuring the comparison was the analysis of the media of commissioner power outlined in Chapter 2. This method contributed to answering RQ2, RQ3 and RQ4.

4. A systematic comparison, using the same framework, of patterns across the English case studies with case studies in Germany and Italy. This method contributed to answering RQ3, RQ4 and RQ5.

5. Action learning sets of commissioners from the English, German and Italian case study sites. This contributed to answering RQ1, RQ2 and RQ3.

6. Framework analyses to synthesise the above findings, contributing to answering all of RQ2, RQ3, RQ4 and RQ5 and to testing the programme theory’s underlying assumptions.

These methods fitted together as follows. The discourse and Leximancer analysis revealed the empirical and causal assumptions (programme theory) on which current commissioning policy rests, especially about how commissioners can influence the providers of NHS services (RQ1). Case studies of the development of local commissioning then explored how far the assumed commissioning organisations and systems were present to begin with (RQ2). The case studies were also used to explore which commissioning mechanisms commissioners were using, how they did so (RQ3), if they used different mechanisms for different care groups (RQ4), how providers responded and what contexts appeared to influence providers’ reactions to commissioning (RQ5). Within the data availability constraints, the cross-sectional analysis of managerial data served the same purposes. Action learning was another way to explore, in ‘real time’, what mechanisms commissioners were using. We used the international comparisons to explore and differentiate which commissioning contexts, mechanisms (above all, media of power) and outcomes appear common to quasi-markets more widely, and which are peculiar to English NHS commissioning. Each method contributed some parts of an overall, perforce incomplete, jigsaw of the complex relationships between commissioning organisations, contexts, mechanisms and outcomes. Table 1 gives an overview.

Because all the above were undertaken during 2010–12, the cross-sectional study data are for 2008–9 (the latest available in 2010–12). Comparison with published managerial data, where available, helped indicate the likely generalisability of our findings. We also compared our own findings with relevant empirical findings emerging from research studies in the Health Reform Evaluation Programme, HSDR, National Institute for Health and Care Excellence (NICE) and the DH Health Policy Programmes.

We involved patient representatives [through PenPIG, the Patient Involvement Group of the SW Peninsula Collaboration for Leadership in Applied Health Research and Care (CLAHRC)], consulting them before research started about the research questions and overall research design (as expressed in the lay summary) and afterwards about our conclusions (expressed in an executive summary). We involved key commissioning stakeholders – clinicians and managers – through the action learning set described below, and will involve them again in the post-project dissemination activities.

Discourse analysis

Cross-sectional analysis of published managerial data

Insofar as data were available, the cross-sectional analysis of published routinely collected managerial data was designed to:

1. provide a sampling frame for the English case studies

2. describe the profile (mix) of commissioning organisation(s)’ practices and resources in the English NHS, allowing categorisation of local health economies in terms of these variables

3. test for associations between health economies’ organisational characteristics (profiles), commissioning practices and published indicators of service outcomes.

The second and third of these also contributed to answering RQ3 and RQ5.

Systematic comparison of case studies

International comparisons

Action learning

The action learning set (see Appendix 8) aimed to:

1. refine our research questions and tools in collaboration with commissioning managers and clinicians, test the face validity of emerging preliminary findings with them and consider dissemination strategies

2. bring together an international cohort of interested clinicians, managers and researchers to explore the potential for learning from commissioning experiences in England, Germany and Italy

3. offer busy managers and clinicians a forum for learning and development reflecting their own personal and organisational agendas rather than one imposed by commissioning ‘experts’, as a catalyst for participants to work on their own challenging commissioning problems, capitalising on the knowledge and experience within the set.

Items 2 and 3 would also help generate recommendations for commissioning practice.

One clinician and one manager from each of the five study sites participated in action learning sets between November 2011 and March 2013. At the outset the participants established their own:

• ground rules (e.g. about confidentiality) for conducting the learning set

• check-in rules

• update frequency

• reporting process, including rules of confidentiality

• check-out rules, including arrangements for subsequent sets.

Three of the researchers were facilitators. At each meeting we also presented emerging research findings to the participants and provided expert input in specific areas (e.g. supply chain management, findings from other research studies on commissioning). One learning set was combined with a mini-conference in which the learning set members exchanged updates and practical ideas about commissioning with informants from the German and Italian study sites. The learning set members agreed conclusions at each meeting, and from the action learning set overall. The anonymised summary findings were included in the evidence synthesis.

Synthesis

By these methods we assembled and analysed the data summarised in Table 6.

As noted, the cross-sectional method used pre-CCG data for a period of 2 years before the case study fieldwork.

Data were synthesised by framework analysis. Conceptually this was equivalent to constructing, for each research question, a data grid in which each row contained data relevant to some aspect of that research question and each column represented a site; and then populating the cells with the relevant data from the case study ‘pithy sentences’, cross-sectional analysis, action learning set, international comparisons and other published studies. Collating data this way provided an immediate means of triangulation, revealing gaps, ambiguities or apparent contradictions in the data, prompting supplementary data collection. By noting common or divergent patterns across cells, we ‘read off’ answers to our research questions and discovered where new categories or concepts were required to accommodate unforeseen empirical findings. 208 We also compared our own findings against those from other studies. The same method was used for cross-country comparison except that the virtual columns now represented countries.

A second framework analysis recollated these data under the standard CMO categories of realistic evaluation, which allowed us to evaluate the empirical validity of the programme theory assumption.

Ethics and research governance

South West Research Ethics Committee approval (reference 09/H0206/50) was subject to informant anonymity and to obtaining management permission from each research site, which we did for all the NHS sites. University of Plymouth ethical approval was obtained for non-NHS sites. Obtaining study site access was laborious and slow (many months) because of a combination of site staff overload (see Chapter 6, section Overload), cumbrous procedures for obtaining the permissions, and administrative delays. The growing importance of evidence-based policy-making contrasted with the difficulty of accessing the necessary evidence. No such difficulty arose in Germany or Italy.

Findings

Our findings are presented by research question, indicating as necessary what their evidential basis is.

Quantitative (Leximancer) content analysis

The Leximancer analysis aimed to list, and quantify the salience of, the assumptions about the structures, mechanisms and outcomes of NHS commissioning stated in our samples of documentary and oral material. Note that in Leximancer analysis the term ‘concept’ denotes only the repeated co-occurrences of terms in the texts analysed. ‘Themes’ are repeated co-occurrences of concepts. Tables 17, 18 and 19 in Appendix 1 show in descending order of frequency what concepts concerning contexts, mechanisms and outcomes were found when the three text samples were combined. For each of the seven outcome-concepts, Table 20 in Appendix 1 shows how frequently each mechanism-concept co-occurred. Appendix 1 also shows the frequencies of occurrence of concepts relating to the outcomes, mechanisms and contexts for each of the three samples of policy text separately.

The mere presence of these associations did not, however, reveal what specific context–mechanism–outcome relationships the policy texts were assuming. The concepts found by Leximancer analysis were too broad for that. We therefore extracted from the source texts each passage in which these associations occurred. Having removed substantively duplicate passages, we made a précis of each that remained, using a formulation suitable for empirical testing; that is, ‘Doing X under conditions C will cause agent A to do Y’ or a logical equivalent. Sometimes these passages described CMO relationships explicitly but more often they were elliptical, calling on the reader’s background knowledge to supply the implicit missing elements. Then we supplied the apparently missing assumptions, which are italicised to distinguish them from the original explicit texts. Appendix 9 shows what we added, and that the additions were few (5% of the programme theory word count).

The analysed materials contained many assertions about what organisations were needed, their duties and defects in existing commissioning systems. They only briefly described how CCGs, the main NHS commissioning mechanism, would work. Apart from mentioning ‘effective dialogue’ with hospitals, almost all that Equity and Excellence: Liberating the NHS says on that point is that clinical commissioning

will bring together responsibility for clinical decisions and for the financial consequences of these decisions . . . reinforce the crucial role that GPs already play in committing NHS resources . . . increase efficiency, by enabling GPs to strip out activities that do not have appreciable benefits (p. 27). 195

Little detail appeared, either, about the mechanisms of patient involvement in care decisions; about how public health measures, research and evidence basing will improve patient outcomes; about how joint working between NHS commissioning groups and local government, patient involvement, research and evidence basing will help contain NHS costs; or about what makes a provider ‘qualified’. Much of the 2012 Act itself was uninformative on these points (naturally focusing on defining terms, public bodies’ powers of appointment, remuneration, accounting arrangements and amendments to previous law),196 but the official Factsheets published to explain it209–228 did describe examples of effective commissioning. The speeches and, much more, Select Committee material229–236 explained more fully some of the mechanisms and mechanism–outcome relationships left unexplained in the written material. In summary, the policy documents and public debates comprised mostly descriptions of contemporary practice and normative assertions about the new organisations’ roles and duties, but offered only a few, mostly meagre, accounts of the mechanisms or the reasoning behind them, some of which may be post hoc rationalisations. 22

Cognitive frame analysis

To reach a more complete, coherent account of the programme theory, it was necessary to impute more explicitly the missing assumptions that policy documents had left the reader to supply. We used our data from interviews with policy-makers and senior implementers to fill these gaps, especially about what Leximancer analysis had shown to be the most salient CMO relationships in the policy texts: competition (patient choice, contestability), financial incentives and negotiations between commissioners and providers.

There was also the question of how other NHS governance structures would interact with commissioning, and how the different commissioning structures would interact.

Commissioning isn’t the route by which improvement is . . . necessarily . . . driven in the system . . . [But by] peer pressure within providers, professional competition, information data and . . . public pressure around that choice, the extent to which punters will be moving, there will be regulatory pressure at the bottom end as well.

Senior manager 1

Another senior civil servant agreed, adding that accountability and audit arrangements would be stronger than under GP fundholding. Transparent, accessible information would enable patients to compare commissioners and choose their hospital. Regulation would guarantee minimum standards and safety of care.

What the [2012 Health] Bill is trying to do is to try and find a, kind of, pragmatic way through the middle that combines the best aspects of all of those different things. I think if you only have integration that can’t be the way forward and if you only have competition, and no integration, then that can be problematic too in certain circumstances. So I think it’s trying to somehow try and steer the middle ground.

Senior manager 2

The new commissioning mechanisms were intended partly to strengthen commissioners’ control over NHS costs:

[C]ommissioners are . . . most of the time . . . merely acting as literally the payers at the end of the day, on stuff that they don’t really have much control over. . . . And the reason why they’re writing the cheques is that there are a bunch of other people in the system, GPs . . . actually incurring the expenditure.

Senior manager 2

Yet commissioners’ role would be narrow:

[T]he only way a provider is accountable to a CCG is via the contract. There isn’t another thing.

Senior manager 3

Hence:

[T]hat doesn’t really get them [CCGs] into the arena of managing providers within a local health system, it’s completely different from that.

Senior manager 4

Policy-makers suggested that population-based, general practice-based and client-based commissioning were compatible and coherent because they were all designed to commission services on the patient’s behalf:

So the main function of commissioning is to be an effective agent on behalf of the people for whom they are purchasing care in ensuring that people get access to the right level of care or sufficient quality in the system in way that actually doesn’t exceed the amount of money that is available.

Senior manager 1

General practitioners will commission as patients’ proxies, because patients will be able to select GPs for, among other things, referral opportunities:

If people in one geographical area can see in comparison to another nearby geographical area that their GPs are delivering a poorer-quality service with less good outcomes, less access to pharmaceutical products, slower referral times, da, da, da, etc. etc. they’re going to want to say ‘Well why? I’m going to move my personal register to another GP practice’.

[. . .] You seem to be envisaging a situation where if the patient moves their registration then presumably some fragment of the GP’s commissioning budget eventually goes?

Yes, the money for the patient, yes.

As for media of commissioner power, Parliamentarian 2 suggested that, within commissioning consortia, managers would perform much of the commissioning work while GPs undertook the overall governance. For both

a starting point [is], I think in terms of them being clear about your commissioning strategy, and particularly priorities for investment and disinvestment, making sure they are coherent, they tie back to need, they’re affordable, but also that you engage with providers at an early stage.

Senior manager 5

He added that commissioners’ negotiations with hospitals were intended to ‘rebalance, empower commissioners’ in negotiating with providers. GPs could not only negotiate with hospital consultants more critically and in greater depth than lay managers could, but thereby involve consultants in managerial decision-making:

So the aim is to actually increase clinician engagement across the pitch but using GP commissioners as the way in.

Senior manager 3

The negotiations would concern technical efficiency (care pathway design) and service cost, and stimulate consultants to raise the quality of, even reconfigure, secondary care. The concomitant risk was that GPs and consultants might collaborate to stymie provider competition. In a second set of negotiations a CCG’s central co-ordinating body would review its own member practices’ quality of care and ‘work with’ individual practices to improve it:

[S]howing clinical leadership to improve services at scale and pace . . . [so that] the whole primary care body start behaving along the lines of the redesigned services.

Senior manager 6

Peer pressure within CCGs would induce GPs to refer as CCGs want. GPs would also have the interest and capability to integrate more fully the now-fragmented general practice, community health, social care and outpatient services.

Transparency was the main discursive mechanism (as we have called it) putting pressure on providers to improve service quality. Commissioners and patients would be able to

monitor performance using a range of mechanisms . . . There’ll be the hard information they receive, the comparative information they’ve got to look at performance compared with other areas. There’ll be the soft intelligence they get through their own patients, through the public and other parties, HealthWatch, CQC [the Care Quality Commission], the Commissioning Board and so on.

Senior manager 4

Commissioners’ asymmetry of information with secondary providers would be less for GPs (as clinicians) than for non-medical managers. Furthermore:

There is of course a political dimension to this, because any politician, any MP who doesn’t stand up for his or her local hospital won’t be the MP for very long. And that’s why it [commissioning] needs to be clinically led rather than politically led, and also politicians need to say when a clinician says ‘well I think we need to do X, Y and Z’, rather than saying ‘oh no, oh no, how terrible’, say ‘well actually that’s right, I agree with that because it’s going to improve patient outcomes and survival rates, experience etc. etc.’

Parliamentarian 1

Tariff payments would give providers financial incentives and the means to self-finance their own investment and/or disinvestment strategies. Also:

[C]ommissioners will have some say over [tariffs] locally, up to a point, and then the national commissioning board have a key role in structuring the tariff.

Senior manager 1

Tariffs would also make providers face the financial consequences of what they did. Commissioners would be able to

say ‘I’m withdrawing my contract from you in six months if it [your performance] hasn’t improved’ and then you actually do it. . . . And then if you could work it out and you had a failure regime where the service didn’t close but the organisation collapsed and moved to another place [sc. new management], you could see driving improvement being accelerated.

Senior manager 6

They would not have to solve providers’ debt and solvency problems, which would help deal with problematic providers and hospital overcapacity (e.g. in London). Also the new financial incentives would raise service quality:

shifting the least good or the lower mean towards the mean or a bit beyond the mean.

Senior manager 1

Only a limited range of standards would be incentivised across the NHS at any one time. The tariff system might also be revised to include ‘year-of-care’ payments for chronic ill-health, because existing tariffs incentivised providers to increase activity.

Competition was intended to raise service quality, reflecting a general policy assumption that markets ‘deliver efficiency and innovation, and improve the quality of services’ (www.oft.gov.uk/shared_oft/economic_research/oft887.pdf), and increase provider diversity:

so that it’s not a monopolistic capture, you can have more than one provider and you can have some dynamism and change and some sense of challenge.

Senior manager 1

New provider entry was a means of introducing new models of care. (Two different informants both cited Diabetes UK providing primary care management of diabetes as an example.) Similarly, the AQP policy would mean that:

for those procedures where you’re able to specify . . . anybody who can meet the requirement can provide the service . . . I suspect it’s more around pathway diversification so seeing the more traditional forms of pathway delivery being changed.

Senior manager 4

Finally, provider competition would facilitate patient choice of provider, a policy goal in itself. Both interviewees and policy documents mentioned that staff-owned providers (social enterprises etc.) might be commissioned, but their formulations suggested only weak interest. Provider competition was national policy, not discretionary, although local commissioners could decide which services to place under the AQP mechanism rather than competitive tendering.

Informants mentioned juridical controls mainly as mechanisms preventing commissioners (and providers) from acting anticompetitively. The European procurement requirements were, senior manager informants said, very general and high level. The UK government (through the 2012 Act196) would in practice decide what specific procurement requirements NHS commissioners faced and not go far beyond those already applying to PCTs. Professional regulations would prevent competition or patient choice having any adverse effects on quality of care:

[A]ny provider organisation, whether it’s a hospital or a community service or a general practice, they know that at the end of the day their duty is to provide the right care for the individual patient and as an individual clinician they could be in breach of professional regulatory standards or as an organisation they could be in breach of contract if they ever did something that . . . was the wrong decision for that patient.

Senior manager 7

Contracts were seen largely as a framework for specifying services. However:

If GPs and hospital doctors don’t agree or are at war or at least can’t see a way ahead, then actually the contract will not solve any of the issues by and large that they’re throwing up.

Senior manager 8

Disagreements or problems about service redesign, cost or quality would normally be resolved negotiatively.

Policy-makers’ and top managers’ programme theory

Combining the explicit and the imputed material, Figure 1 represents the overall structure of the programme theory of NHS commissioning policy during 2010–12. It combines what have been called ‘clinical’, ‘managerial’ and ‘political’ discourses. 237 Arrows represent the main mechanism–outcome relationships that the policies assumed. Figure 1 represents high-level mechanisms rather than all the intermediate processes nested within them, such as GPs’ and patients’ everyday care choices. Each concrete mechanism noted in Figure 1 may combine several media of power. ‘Providers’ include community health, mental health and non-GP services besides hospitals.

FIGURE 1.

The underlying programme theory: NHS commissioning policy 2010–12. LA, local authority; NCB, National Children’s Bureau; OI, outcome indicator; PH, public health; PHR, primary health care.

Each arrow in Figure 1 denotes a testable proposition or complex of them. Formulated as CMO propositions, the main programme theory assumptions of NHS commissioning policy were:

PT1: Central government’s decisions and actions lead NHS Commissioning Board, NICE, etc, to set regulations and guidelines that reflect ministers’ priorities.

PT2: Through joint commissioning with local government, commissioners will understand care group and population health needs.

PT3: NHS England’s management and GPs’ knowledge of patient needs together allow and stimulate GP commissioners to link clinical and financial decisions.

PT4: General practitioners’ commissioning role will lead CCGs to review primary care quality.

PT5: General practitioners’ commissioning role and joint commissioning with local government will lead to stronger integration of primary health and social care services (before and after hospital episodes).

PT6: General practitioners’ commissioning decisions, reviews of primary care quality and stronger partnership working with social care will bring about reconfiguration of primary care pathways, removing ineffective activities.

PT7: The reshaping of primary care pathways and the development of new information technology (IT) (online consultations etc.) together result in primary replacing secondary care to some extent.

PT8: General practitioners’ commissioning decisions, the outcomes framework, the presence of alternative providers and patients’ choice of providers (in particular of GPs) together stimulate and enable GPs to exercise greater influence over other providers.

PT9: Increased GP influence over other providers and the substitution of primary for secondary care together improve clinical outcomes.

PT10: Increased GP influence over other providers, GPs’ focus on quality and the substitution of primary for secondary care together improve quality of care.

PT11: Increased GP influence over other providers and the substitution of primary for secondary care together reduce costs of care.

PT12: Ministerial decisions lead NICE to produce the specific outcomes framework.

PT13: The outcomes framework and IT developments result in patients, commissioners and providers themselves having information by which to compare providers’ performance, and in patients and commissioners selecting (and deselecting) providers.

PT14: The outcomes framework, information for patients, provider plurality and (later) personal budgets lead patients and commissioners to choose between (and deselect) providers.

PT15: Commissioners and patients choosing between health-care providers causes the latter to improve clinical outcomes together.

PT16: Commissioners and patients choosing between health-care providers causes the latter to improve quality of care.

PT17: Patients choosing between health-care providers causes the latter to reduce the cost of care.

PT18: Research and evidence production causes health-care providers to improve clinical outcomes.

PT19: Patient involvement in care decisions causes health-care providers to improve clinical outcomes.

PT20: Public health activity causes health-care providers to improve clinical outcomes.

PT21: Refusing ‘bail-outs’ causes health-care providers to reduce costs.

PT22: General practitioner and local government joint commissioning causes health-care providers to reduce costs.

PT23: Patient involvement in care decisions causes health-care providers to reduce costs.

PT24: Research and evidence cause health-care providers to reduce costs.

PT25: The availability of information by which to compare providers’ – and by implication commissioners’ – performance increases NHS organisations’ accountability to the public.

These relationships are moderated by the presence or character of GP engagement with commissioning, local experience of practice-based commissioning, patients’ willingness to choose their hospital and on that basis their GP if necessary, the availability of ‘qualified’ competing providers, and existing providers’ capacity and scope to respond to financial incentives and competitive pressures.

The above propositions, and Figure 1, omit many details, nuances and complications, outlining only the gross overall architecture of the exposed programme theory. As with previous policies,22 the policy aims were broad and difficult to pin down. Nevertheless, the above propositions are amenable to empirical testing in light of NHS commissioning practice, to which we now turn.

Reconfiguration as recycling

In all our study sites CCGs evolved from existing organisational structures for GP involvement, but which ones existed, were used and reconfigured varied between sites. In Porttown the starting point was a social enterprise (community interest company), which originated as the practice-based commissioning coordinating body, concentrating on referral management and owned by the city’s GPs. After the 2010 election the PCT transferred the Professional Executive Committee’s (PEC) commissioning functions to it. As its role extended, the social enterprise was ‘rolled up’ into the CCG. Paradoxically, the 2012 Health Act was an impediment. 196 To work around its requirement that the CCG be a statutory body, the PCT created an interim ‘Clinical Commissioning Executive’ to work after the fashion of a holding company. In Milltown those parts of the PCT in which GPs were active (the PEC, various micro-commissioning groups) evolved into the CCG by ‘scaling up’ already-established arrangements. In neither site had the PCT been clustered with another. Metroland split the PCT cluster (and other cross-London groupings) back into four, creating one CCG for each borough in the cluster’s territory. It populated its CCG from an existing polyclinic (or ‘polysystem’):

[S]o they’ve actually basically kept that poly, the sort of, there were, sort of, four, five, if you like groupings within the PCT and they’ve basically kept that within the CCGs.

General practitioner, Metroland

General practitioners involved in the Metroland polysystem became CCG enthusiasts, but others were generally unfamiliar with commissioning. Consequently the new CCG was constructed almost from scratch, initially with six GPs and designated PCT staff to assist them. Shire formed its CCG by clustering two PCTs. One had developed autonomous locality structures that undertook practice-based commissioning. Clinical commissioning programmes were long established there, with PCT, GP, other primary care and local authority representatives meeting acute and mental health trust medical directors every month. This PCT became the platform for CCG formation, absorbing the existing cardiac, stroke and diabetes networks. The other PCT had little practice-based commissioning and had to start recruiting GPs into commissioning work with the other PCT almost from scratch.

The transition to CCGs blurred the division of commissioning labour between PCT, cluster and shadow CCGs, which retarded commissioning work in all four sites. Metroland informants were also uncertain whether the CCG should be shaping or implementing GPs’ wishes. Those in Shire were uncertain whether the CCG should focus on primary or secondary care.

Incremental engagement

Another common pattern across sites was difficulty in recruiting GPs to ‘engage’ with commissioning. PCT informants in Metroland, Shire and (more guardedly) Porttown emphasised how fast this was happening, but some GPs presented a more nuanced view.

Wider collaboration

A parallel task was to reconfigure joint commissioning with local government and GP negotiations with hospital, mental health and community health services trusts.

Overload

These developments occurred despite the disruptive side effects of NHS restructuring on staff turnover (although the disruption was least in Metroland). Three successive changes of Shire commissioning structures meant that:

They [commissioners] never develop that expertise that you can start having a sensible dialogue.

Medical director, Shire

It caused

loss of organisational memory.

General practitioner, Metroland

Staff turnover and increased work demands forced those remaining to prioritise tasks. Metroland and Porttown PCT rotated their attention around care groups so that some service specifications were reviewed in depth with ‘just programme managing essentially’ (manager, Porttown) for the rest. Shire ‘rationalised’ 22 Clinical Commissioning Programmes into 7, aligning each with a national programme budget heading, but also making agendas and meetings (20–40 people) too large for easy decision-making, and mostly irrelevant to the hospitals. This made clinicians see the groups as mostly for information gathering. Other tasks were simply delayed, for instance out-posting hospital diabetic nurses to general practices (to reduce outpatient attendances; Shire) or finalising plans for social work assessments in winter of older people with long-term conditions (Metroland). Commissioners also engaged external commissioning support: management consultants (e.g. to review the viability of rehabilitation beds; Metroland), the London Health Programme (Metroland), the Procurement Supplies Agency (Porttown), the SHA (Shire, Milltown) and independent experts (Shire). Livewell bought substantial support from a US HMO to develop a unified view, across the commissioners, GPs and other primary care providers, about what services should be commissioned, monitored and reviewed. 238 Overload was less evident in Milltown, perhaps because so many GPs contributed to commissioning there. It became more severe the year before CCGs went live, with less work being done on commissioning initiatives, revising service specifications (Metroland) and setting public health priorities (Porttown).

Convergent reconfiguration

Despite their different origins, the study CCGs converged on a common structure, partly stipulated by legislation and central guidance: clinical senates; subgroups for specific care groups, clinical specialities or service problems; and the shadow CCG that from April 2013 would manage the first two. Commissioners created their organisational structures partly in order to involve clinicians (e.g. the 42 project groups in Milltown).

Recycling existing organisations implied recycling the existing local GP leaderships, as others have also found. 239 The Shire GPs dominating service redesign work and chairing clinical commissioning programmes tended to be former PEC members, GPs with a special interest (GPSIs) or, as in Porttown and Milltown, former practice-based commissioning activists. The former PCT chair in Metroland reappeared as assistant director of the CCG. As noted above, Milltown PCT took steps to moderate, but not reverse, this tendency. These continuities of GP leadership, of subgroups doing much of the commissioning, and of non-GP managers doing the commissioning work that GPs would not, seemed to presage more an evolution of commissioning practice than the upheaval, even ‘revolution’, that many commentators and policy-makers predicted before April 2013.

Management performance

Separating commissioner and provider forced commissioners to be more deliberate, focused and explicit (documented) in specifying what services they wanted from Porttown and Milltown community health services. Existing providers still won the tender, but had to specify services and costs more precisely.

Service specifications were becoming evidence based, as others have found. 102,240 In all four sites, public health professionals contributed local data on which to base service specifications. Shire PCT did so least, but (like Porttown) did use the Map of Medicine (http://mapofmedicine.com) to redesign care pathways and specify referral criteria. Porttown and Metroland also re-used Hospital Episode Statistics data. Their public health staff contributed to service specifications and pathway design. When evidence was absent or ambivalent, medical consensus or majority opinion was followed – again reported elsewhere102 – and commissioners made more subjective judgements (‘intelligence’) based on financial information, their own perceptions of local services, and staff and patient feedback (e.g. complaints).

Providers, indeed the main provider, played a large part in formulating service specifications. Commissioners in all sites spoke of provider proposals for service provision, which ranged from consultants advising about the evidence base for service specifications (Porttown) to consultants telling commissioners what service changes were feasible (Metroland). Provider input compensated for the commissioners’ lack of clinical and technical knowledge, as provider informants often commented. Nevertheless, mental health commissioners sometimes argued that:

I think that I can commission better because I was involved in providing the service.

Mental health commissioner, Metroland

Clearer service specifications helped in introducing referral-screening criteria aimed at reducing hospital activity by diverting suitable patients to primary or community care, at least for extra-contractual referrals (Metroland). Where referral criteria were lacking, peer review was used, often with consultant (Porttown, Metroland, Shire), and in Shire also physiotherapist, input.

Commissioners in our study sites often felt that they lacked the procurement skills required for dealing with competing providers. They were aware, however, of the transaction costs of procurement work, which they saw as a somewhat futile waste of money and time when only one bidder was available. Informants at Shire contrasted the local authority’s ‘scrupulous’ application of tendering rules with the small number of providers. Our study sites followed official procurement procedures and regulations most carefully for large contracts and when there were two or more bidders, but, if market testing identified only one potential provider and they could keep within EU law, they would avoid tendering.

Commissioners mainly used nationally mandated measures such as target compliance, patient-reported outcome measures (PROMs)241 and DH ‘vital signs’ for monitoring providers. During our fieldwork the latest NHS Operating Framework242 measured recorded quality, resources and reform. Two (of eight) of the headline and nine (of 27) supporting quality measures concerned health outcomes; the rest measured work processes. (No resource or reforms targets concerned health outcomes.) The NHS Outcomes Framework had not yet appeared. 243 Porttown PCT benchmarked its providers against regional comparators, as did Shire. All sites used contract reports that included Monitor’s reporting requirements, QIPP data and exception reports for non-compliance with contract or quality standards. Metroland required local providers to give ‘narrative’ explanations of outliers, exceptions and untoward incidents. Metroland, Shire and Porttown encouraged provider staff to report service quality problems. Porttown commissioned a local firm to undertake patient surveys, but dealt directly with any patient who wrote comments (rather than just scores) on the survey form. Livewell collected data via its telephone helpline.

Although commissioners relied on their providers to supply and interpret monitoring data, providers sometimes withheld information, such as ‘commercially sensitive’ financial data, on the grounds that they had to make a surplus (Metroland). Non-disclosure could have adverse consequences for both sides. Porttown hospital trust masked its financial problems as long as it could until the PCT and hospital found themselves having to make savings exceeding £30M. Tariff payments made provider finances, patient flows and case mix more transparent than did block contracting. Similarly, because there were QOF payments for diabetes and epilepsy treatments, Shire PCT knew how many of those patients were treated, but knew the number of Parkinson’s disease treatments only because it employed a specialist nurse. At Metroland especially, our informants thought that involving GPs made service monitoring more incisive, especially regarding clinical outcomes and cost-effectiveness.

The action learning participants suggested that an important aspect of management performance (especially service specification and contracting) was to promote integration of the disparate primary, secondary and social care pathways that an individual patient followed, although, in practice, commissioning was rarely carried out in this way. Often, discussions between commissioners and providers about how to identify patients with long-term conditions who might be diverted to community care (to reduce costs) reached an impasse. This was an obstacle to reducing inpatient activity.

Looking at all English PCTs cross-sectionally, we had data on 16 service outcomes (see Chapter 4, section Indicators and measures). Only one, PCT surplus/deficit, correlated with PCTs’ summary WCC scores (see Table 23, Appendix 4), and then weakly and in the opposite direction from predicted; larger PCT deficits (or smaller surpluses) were associated with higher WCC scores. However, PCT surplus also correlated negatively (r = –0.319, p < 0.001) with monthly mean waiting list size (combined inpatient and day cases) per 100,000 population: the shorter the waiting list proportionate to population, the smaller the PCT’s financial surplus (or the larger its deficit). The more numerous correlations between the controls and the service outcomes were, with two exceptions, also verging on negligible. One might have expected WCC scores to reflect among other things the implementation of contemporary managerial objectives. 55 If so, our finding casts doubt on the validity of the WCC self-scoring system (which was subsequently discontinued).

During the study period the Francis Report244 exposed a culture of vague monitoring criteria and unrealistically optimistic findings. Francis reported that commissioners (and providers) often responded to prima facie adverse monitoring reports by reinterpreting the findings rather than acting on them. (Stafford was not one of our study sites.)

Negotiated order

Contract negotiations were conducted mostly by chief executive dealing with chief executive, and other non-medical managers dealing with their counterparts (e.g. PCT finance manager with trust finance manager etc.). Senior clinicians sometimes participated, but not uniformly, especially not in negotiations about cost savings. In Metroland mental health trust, even the medical directors were not always involved in such negotiations. Negotiations about the unscheduled care contracts in Porttown involved the PCT and acute trust finance directors, with marginal input from other PCT commissioning staff, let alone anyone else:

It’s just a contractual numbers negotiation.

Commissioning manager, Porttown

Shire was similar. There clinicians’ proposals (e.g. for diabetes pathways) would not necessarily enter contract negotiations. One Metroland consultant believed that ‘difficult’ consultants were kept out of contract negotiations (no one else either corroborated or denied this). One way to bridge the two negotiated orders was by judiciously selecting consultants – and GPs – to involve. Porttown hospital trust involved selected consultants in order to demonstrate to them (and, through them, other consultants) that the trust had obtained the best deal it could. Action learning participants reported instances when a consultant spent much time negotiating the introduction of new services and reconfiguring care pathways with one commissioning manager, only for the agreement to be reversed elsewhere in the PCT. The consultant then felt she had wasted her time and spent less time working with the commissioner.

National policy priorities (targets, budgets) largely set the commissioners’ agenda, as another study102 also found.

Commissioners need to demonstrate that they exercise control. They are only interested in national diktats and money.

Consultant, Metroland

However, the national policy focus did give providers and commissioners common interests, especially concerning hospital referral criteria, substitution of primary for secondary care, compliance with QIPP targets, emergency department (ED) utilisation and hospital outcomes, although Porttown community health service (CHS) managers argued that their commissioners paid too much attention to minutiae (e.g. heart failure nurses’ referral habits). Indeed, our hospital informants described the main commissioning focus as being costs, although in Metroland waiting times were no less central. Action learning set members used the term ‘class hegemony’ to describe the close networks among small numbers of nationally prominent NHS managers, networks which local managers would use when trying to influence the other party in commissioning negotiations.

Financial negotiations were often separate from ‘real-side’ negotiations. In Shire, Metroland and Milltown, a financial agreement (and, in Milltown, also local authority budgets) framed subsequent ‘real-side’ negotiations about service priorities, population needs and service specifications. In Porttown, for instance, the clinical and service specification of a new eating disorder service was adjusted to meet the budget, not vice versa. The main exception was that, when the London Health Programmes awarded contracts for ‘hyper-acute’ trauma and stroke centres, specifications for quick access and clinical quality determined the cost. As others report,102 both financial and ‘real-side’ negotiations were framed by historical patterns:

[W]e commission most things on a historic basis unless there’s a good reason to start arguing and we need to change things . . . you’re basically arguing around the margins with the provider about what changes. But . . . where there are significant risks around delivery . . . you’re likely to take more of [a] care pathway approach to such as A&E [accident and emergency], such as maternity.

General practitioner, Metroland

Commissioners’ proposals for cost containment and service re-profiling were bound to be contentious. Providers who disliked commissioners’ proposals disputed the justifications for them. Shire PCT and one hospital argued inconclusively for years about why ED attendances were increasing. ED consultants blamed the lack of GP availability; the PCT cited its own data and surveys suggesting the opposite.

A second negotiated order involved GPs and consultants (and other secondary care clinicians), who often liked to deal directly with each other, blurring the commissioner–provider distinction (especially in the small-scale Livewell project) and bypassing the aforementioned managerial negotiations. Clinicians’ attitudes towards non-medical commissioners ranged from:

We don’t just want to be a manager with a clinician’s hat on

General practitioner, Porttown

to:

I do my level best not to deal with commissioners.

Consultant, Metroland

Livewell GPs disliked even the word ‘commissioning’. 238 Doctor-to-doctor negotiations could have constructive results for a commissioner. Porttown consultants endorsed changes to ear, nose and throat services as a consequence of dealing directly with the GP organisation, while the GPs questioned and reportedly limited the orthopaedic consultants’ income-maximising activities. The obverse of this medical collegiality was, at times, scepticism about the value of non-medical inputs to commissioning and about managers trying to control things in which they were not competent. There were also limits to medical collegiality:

[O]ne of the hardest things I’ve found is convincing my GP colleagues that the hospital doesn’t have a hidden agenda, it’s not out there to ruin general practice.

General practitioner, Milltown

Porttown and Milltown informants described mediating debates between GPs and consultants. Disagreements about what work should transfer to non-hospital settings soured relationships between Shire GPs and consultants.

In terms of quantity of work and participants, the most important negotiated order was what GPs called ‘micro-commissioning’: interdisciplinary, interorganisational redesign of care pathways, most often (and protractedly) for complex, non-commodified activities such as rehabilitation, chronic obstructive pulmonary disease (COPD), unscheduled care and mental health care. However, care pathways for less organisationally complex care – for example orthopaedic surgery – were also sometimes micro-commissioned. Micro-commissioning typically involved several providers, NHS and local authority commissioners, patient representatives and ‘third-party’ experts such as a lead psychiatrist, public health consultant (Metroland) or immunologist (Porttown), but could also be small scale (e.g. at Livewell). From the pathway design followed decisions about providers’ functions, service access criteria, resource needs – including decommissioning decisions in Milltown – and in Livewell an overall redesign of primary care provision. From these decisions followed service specifications and monitoring arrangements, raw material from which contracts were prepared. Micro-commissioning gave the participating clinicians and providers a head start in bidding for the ensuing contract.

Building up trust with providers took years. It required compromises and trade-offs (as others report39), and partly depended on the personalities and interests involved. After Milltown PCT decommissioned its community matron service, it collaborated with (other) providers to redeploy the community matrons, which the commissioner saw as investing in future goodwill. Metroland PCT agreed that its hospitals could cut beds (hence costs) provided their activity was not reduced, and paid for some (not all) of the above-contract activity that the acute hospital had undertaken. To resolve a potential dispute about upcoding, a private hospital invited Metroland PCT to send a coder to agree the disputed codes jointly with its own coders. Shire PCT claimed to avoid disputes by mutually recognising divergent interests and reaching compromises. As others report,39 regional NHS bodies (SHAs, during most of 2010–12) would sometimes intervene in contract negotiations to exercise hierarchical control over both sides, including dispute resolution. Shared beliefs in a commonality of NHS interests and in NHS-wide interorganisational collaboration, persisting from earlier NHS structures and rules, also helped mitigate tensions and disputes, indeed facilitated micro-commissioning.

In all four sites, patient and carer representatives (e.g. from Diabetes UK) were involved in micro-commissioning and, in Shire, decisions about service closures. Patient voices were not necessarily unanimous. Some patient groups in Shire opposed acute bed closures, while others preferred replacing acute beds with care closer to home. User representatives for Milltown mental health services shifted over years from an adversarial to a more supportive stance. Porttown patients whom the PCT consulted were interested in wheelchair supply but not vascular risk assessment or ISTC orthopaedic services. For mental health, one challenge was to engage with users at the more disabled end of the spectrum, for instance those with chronic schizophrenia. Neither did patient representatives merely respond to NHS initiatives. Porttown patients participating in a regional heart/stroke network (as opposed to the PCT patient consultations) presented their own proposals for a CHD self-help service. Porttown council had learnt that user involvement worked best when users participated in the early stages of service design and specification.

Over time, negotiations usually became more relational as trust and goodwill accumulated, the participants recognising that they would need each other’s help in future. In Milltown and Porttown, GP–consultant relationships rested on the precept ‘If you have got a problem, ring me’, forestalling the use of other media of power. This social capital was reported in Milltown as helping to repair the harm that tariff payments did to commissioner–provider relationships. Metroland also built on the relationships formed in operating its ‘polyclinic’. Because they influenced what media of power a manager was usually inclined to use, certain personal attributes, skills and management style were important for commissioners, as others have also found. 245 One such skill was motivational interviewing (e.g. inviting providers to state their view of the evidence base to frame negotiations in terms that the provider accepts). Another was to blame a third party (e.g. national policy, the behaviour of another local organisation), not the other negotiator, for the dispute. Patience and emotional resilience were two more. 238 A corporate hospital manager explained:

I’m from a sales background so you learn never to burn a good relationship because it costs you down the line.

Private hospital manager, Metroland

Nevertheless, some relationships had already been burned by personality clashes (Milltown), argumentative, competitive and antagonistic negotiating styles (Shire) or bullying:

I’ve been in meetings that were so bad that if you’d filmed them you’d be in front of a court.

General practitioner, Porttown

Indeed, action learning participants suggested that our analytic schema should include bullying as a method which negotiators used.

Discursive control

Etic discourse framed much of the micro-commissioning negotiation mentioned above and was a main currency of provider monitoring. Thus Shire PCT stated that services without an evidence base would be decommissioned. Commissioners in our study sites mainly used nationally promulgated materials when persuading providers to accept commissioners’ proposals, most often NICE guidelines, the Recovery Model, QIPP targets, the Map of Medicine, National Service Frameworks (especially for mental health, older people and CHD) and the Improving Access to Psychological Therapies (IAPT) programme. Before negotiation or monitoring meetings with providers, commissioners would also supply data on local problems to frame the discussion, for example case-load and case mix spreadsheets (Porttown) or studies showing that a walk-in centre raised ED admissions (Milltown).

At times etic persuasion resolved disagreements between commissioner and provider (e.g. about remodelling eating disorder services in Porttown), but not always. Etic arguments could cut both ways, as they did against commissioner proposals for GP triage in EDs (Metroland) or when the Shire mental health provider simply dismissed national guidance about home treatment as ‘out of date’. ‘Evidence’ was sometimes used to justify rather than ground decisions.

Then debate fell back on emic discourse. Commissioners sometimes cited patient consultation (but seldom patient choice) to legitimise their claims on providers. Porttown PCT asked its providers to show evidence of patient – and staff – consultation about proposed service changes. Occasionally, wider ethical imperatives were invoked. The Porttown pathway group for non-psychotic mental health problems succeeded in remodelling services because everyone recognised the fatal consequences when these services fail. ED consultants in Shire argued that, although they could not control which patients presented at the ED, they could not clinically or morally justify sending patients elsewhere either. The argument that proposed service changes would have no bad consequences for patients was often persuasive to provider clinicians. One influential emic argument reflected the relationality mentioned above, expressing a culture of joint responsibility for the local NHS: ‘the [Milltown] way’. Porttown GPs’ leaders said that their GPs had more appetite for collaboration and integrating their work than for competition.

Financial incentives

There was little disagreement that tariff payments incentivised providers to increase the number of patient episodes, which had

driven a wedge between us over the last few years.

Commissioning manager, Milltown

Commissioners asserted that treatment thresholds had fallen and upcoding had appeared. The volume of providers’ activity became the focus of contract negotiations rather than service specifications. However, rising marginal costs of care limited the expansion. The main Porttown hospital had to open wards and theatres at weekends, at costs above the tariff payments. Shire hospitals wanted to discontinue costly out-posted clinics in community hospitals. Tariff payments constrained micro-commissioning by placing a ceiling on the cost of any new care pathway in Metroland.

Unless commissioners responded as reported below, the combination of tariff system, AQP and patient choice rules weakened their power to control provider case load and case mix, hence overall costs, and the sources of service provision. When Metroland PCT and its main hospital agreed a reduced volume of elective activity, the hospital still deliberately delivered more services than agreed, presenting the PCT with a fait accompli and corresponding bill. Shire PCT found its obligation to pay for hospital ‘overperformance’ an obstacle to transferring funds into long-term care. Porttown commissioner and provider informants agreed that tariffs incentivised hospitals not to transfer services to primary care.

Under tightening financial constraints, commissioners in the study sites responded by bundling their tariff payments to providers into blocks (see also Petsoulas et al. 39), the ‘cap-and-collar’ or ‘managed PBR’ system. If provider activity fell below a certain ‘floor’ or (more likely) exceeded a certain ‘ceiling’ (in Metroland, 5% above the expected volume), tariff payments for the marginal activity were reduced by an agreed amount, eventually falling to zero in Porttown. One Shire hospital, though, so far exceeded its expected volume that the PCT exercised its option to pay it on a ‘non-contractual activity’ basis, that is to pay a monthly sum that was retrospectively adjusted for deviations from the level of activity that the PCT had forecast. There were exceptions; Porttown PCT changed its de facto block contract with the ISTC to tariff payments. Nevertheless, the predominant response was bundling in order to dampen what, according to commissioners, were perverse incentive effects.

Block payments were still used for community health and most mental health services. They sometimes prevented providers from being incentivised, even reimbursed, for work. Metroland mental health trust was not commissioned to treat, say, schizophrenics, nor the Shire trust people for autism, but both trusts were still expected to treat them under a block contract, so that the providers either cross-subsidised the work or sought voluntary-sector provision and/or finance. When Porttown GPs began performing minor operations, the local hospital, not the GPs, was still paid under the hospital’s block contract. Not that block contracts inherently denied commissioners scope to incentivise their providers. A block contract between a Metroland provider and PCT shared the costs of referrals above a specified level in return for the provider increasing its bed capacity. Shire PCT used a block contract to offer its community and mental health service a secure income in return for the providers taking an increased case load. Milltown PCT built into its ED block contract an incentive to send suitable patients to an ambulatory care centre, where the average case cost was £70, not £2000. In theory, CQUIN could also be used as a supplementary incentive mechanism, but the targets and payments were in practice decided nationally and regionally. Within the national NGMS contract, Porttown and Shire used QOF payments to incentivise general practices to implement vascular disease checking. Although in practice the study site commissioners did little to measure the volume of work of block contract holders, well-designed block contracts could accommodate strong monitoring arrangements. One in Shire stipulated the quarterly provision of monitoring information, which was used to review and improve performance, with the commissioner having the option of contract termination.

Commissioners also used block payments to hypothecate funding for nationally prioritised services (e.g. early intervention, Metroland). Conversely, block funding made it easy for Metroland to divert £5M from its mental health provider to meet hospital overspend resulting from tariff incentives, although such transfers also occurred in pre-tariff days. When commissioner budgets were tight, the use of block payments compelled providers to negotiate adjustments with the commissioner; but by the same token Milltown PCT found that this tended to lock the commissioner into using existing providers.

Independent-sector treatment centres were initially paid irrespective of what volume of cases they treated246 and so the Metroland ISTC seemed to feel no incentive to engage with GPs and, we were told, ‘cherry-picked’ patients (PCT commissioning manager, Metroland). That corporate provider, at least, seemed to respond to block payments differently from NHS providers, who were more willing to waive their contractual claims on commissioners.

Like others,71,247 we found that our study site commissioners shrank from financially destabilising the main local providers on whom they depended for fulfilling their commissioning duties. Shire PCT did not exercise its option of non-payment when local providers did not comply with their contract. Neither would Shire or Metroland PCTs reduce total tariff payments to reflect a large reduction in admissions. A Milltown commissioning manager argued that it would be short-sighted to disrupt long-standing, usually productive relationships with a provider for the sake of a one-off financial gain. For voluntary and corporate providers, losing an NHS contract was undesirable but not fatal (Shire, Metroland).

The shift towards lower, ‘best-practice’ tariff prices began only at the end of the study period, as did the introduction of mental health tariffs and (in some places but not our study sites) experimental ‘year-of-care’ (in effect, capitation) payments. In our two study sites where they had been trialled, personal health budgets had so far had little perceptible effect:

I don’t know anything about them.

Commissioner, Metroland

Shire PCT was still ‘investigating the use’ of personal health budgets for brain injury patients (PCT commissioning manager, Metroland).

Provider competition

Provider competition was the medium of power about which the most cross-sectional data were available. Preliminary examination (see Table 21, Appendix 4) showed that spending on independent-sector providers correlated negatively with spending on local authority and voluntary-sector providers, consistent with the interpretation that some PCTs used local authority and voluntary-sector provision, and others used independent-sector provision, in response to hospital overload; and that these two options were to some extent alternatives.

Given our data, the theoretical maximum numbers of correlations, with the sign as predicted, that might be found between measures of competition and of policy outcomes were:

1. provider competition, generically: 1 measure (HHI) × 16 service outcomes = 16 possible correlations

2. population-based competitive bidding: 4 measures × 16 service outcomes = 64 possible correlations.

Ignoring negligibly small ones, 3 of the 16 possible predicted correlations between HHI and the selected service outcomes were found (see Table 24, Appendix 7). HHI correlated as predicted with:

1. amenable mortality rate, all causes, under-75-year-olds, directly standardised

2. monthly mean waiting list, in-patient and day cases, proportionate to weighted population (but this correlation was weak)

3. finished consultant episodes (FCEs) proportionate to weighted population.

Twelve of the other 13 predicted correlations between HHI and the policy-relevant service outcomes were not found. The remaining correlation, between HHI and percentage change in the proportion of all non-admitted patients waiting less than 18 weeks for planned treatment, was in the opposite direction from predicted. The following service outcomes were more strongly associated with a control variable, the PCT Deprivation Index, than with HHI:

1. change in emergency admission for primary care preventable chronic conditions

2. change in emergency admission for primary care preventable acute conditions

3. ratio of observed to expected emergency admissions for conditions not usually requiring hospital admission

4. mean time waited

5. ratio of day cases to admissions

6. average length of stay.

Primary care trust income per capita was more strongly associated with PCT budgetary surplus or deficit than was HHI.

Conceivably, though, gross patterns of association between a generic measure of competition (HHI) and the service outcomes might mask more nuanced associations. We therefore repeated the analysis, replacing HHI in turn with each of the four variables indicating the extent of possible provider competition under population-based commissioning. Of 64 possible correlations, we found four, all with the opposite sign from what the programme theory assumed (see Table 25, Appendix 4). Since the other possible associations were either negligible or absent, it appears that, under population-based commissioning, less concentrated provision of secondary care was not associated with ‘better’ performance on the service outcomes that we considered. Again, control variables (especially deprivation) were more often associated with the service outcomes than were those reflecting the extent of provider competition.

This evidence therefore seems mostly against the assumption that provider competition has helped PCTs realise the service outcomes studied. Advocates of competition might reply that these findings occur because NHS organisations are half-hearted about competition, or because a single provider dominates most local NHS health economies. We therefore repeated the analyses for PCTs with the lowest supply-side market concentration only. For PCTs with the lowest quartile of HHI scores, two more correlations with the predicted sign emerged but the previously found correlation between HHI and FCEs proportionate to weighted population disappeared. Thus 4 of the 16 theoretically possible correlations were as predicted (see Table 27, Appendix 4). The findings suggested that less concentrated hospital provision may be associated with ‘improved’ (reduced) unplanned admissions for conditions sensitive to primary and to ambulatory care; but to complicate matters these patterns probably reflect primary care activity besides hospital competition. We also examined whether the correlation between HHI and waiting lists might reflect PCTs’ use of independent-sector hospitals. For the quartile of PCTs with lowest HHI scores, the correlations between the same service outcome measures and PCT spending on independent-sector providers had the opposite sign from that predicted by the programme theory for mean time waited, for average number of patients waiting more than 4 weeks for a first outpatient appointment (proportionate to weighted population) and for inpatient and day-case waiting list (proportionate to population; see Table 28, Appendix 4). These findings appear to suggest, if anything, that PCTs facing the greatest waiting-list challenges responded by buying independent-sector services. However, PCT spending on independent-sector providers did correlate (with the predicted sign) with the improvement compared with the previous year in:

1. percentages of unplanned hospitalisation for chronic ambulatory care sensitive conditions

2. primary care preventable acute conditions

3. emergency readmissions within 28 days of discharge from hospital

4. admitted patients waiting less than 18 weeks from referral to treatment

5. non-admitted patients.

The findings are also consistent with the hypothesis that PCTs that faced the greatest waiting-list challenges were also those where recently discharged patients and those with primary care preventable conditions were less likely to be admitted to hospital via ED. The contrast between these findings and those regarding HHI reflects the non-correlation between HHI and spending on independent-sector providers (r = –0.0508, p = 0.534). Thus 5 of the 16 possible predicted correlations between spending on independent-sector providers and the service outcome measures were found. Again, the service outcomes studied correlated more often with the controls than with the measures of competition or the use of independent providers.

While the above analyses included the four most obvious possible confounders of the above relationships between provider competition and service outcomes, further research may reveal others. These controversial subjects sometimes provoke objections248 that the researchers selected outcomes likely to produce findings supporting their policy preferences. We minimised that risk by analysing multiple outcomes, and some of the outcome measures we used were sensitive to competition; but since our analysis was limited to published data it remains conceivable that yet further outcomes would be more sensitive to competition. Analyses of the relationships between competition and quality are sensitive to the way market size and London factors are entered into the analysis. 163 The present analysis may corroborate that finding: it defines market size as PCT size and did not adjust for London factors. More important, things may have changed since 2009. For one thing, the policy (in 2009) of regarding NHS trusts as preferred providers was supplanted in 2010 by the Any Willing/Qualified Provider policy. Further research to test whether or not that is so will be required when later data are published.

Our case studies did find some evidence of commissioners using provider contestability as a means of control. Porttown PCT opened negotiations with private nursing companies in case the community health services trust proved unwilling to meet the PCT’s service specification, and decommissioned its wheelchair provider. Metroland commissioners used similar threats to reduce the cost of care packages by 20%. Shire PCT found itself new providers for talking therapies, diagnostics and primary care ophthalmology services. However, Milltown used only the alternative secondary care providers (i.e. the ISTC) imposed on them by higher-level NHS management. Like others,249 we found that provider competition occurred mostly for specific care groups and among small community health service providers, although not, in our study sites, only at their geographical margins.

Existing providers were often the only credible bidder. PCTs had usually (including all our study sites) transferred their community health services en bloc either to the existing, near-monopoly mental health or hospital trust, or to a single dedicated NHS trust or a single social enterprise (Porttown, Surrey). The selection criteria in the London Health Programme’s ‘competition’ for hyper-acute stroke centre contracts included having certain numbers of suitably qualified consultants already available. Pharmacy provision of vascular risk checks was ruled out a priori in Porttown. In Shire the criteria for provider selection included provider experience and relationship with the local health community. Sometimes the third sector also contained only one credible bidder: for Shire memory services, the Alzheimer’s Society.

Metroland and Shire allocated emergency work in certain specialities to one hospital and elective work to another. Whether or not a cartel was intended, there were also volume, hence quality and safety, arguments for concentrating these clinical activities. Shire hospitals, we were told, also had informal agreements about not ‘poaching’ each other’s orthopaedic work. London had 10 mental health trusts but they did not compete. In Milltown, services designated as ‘pilot projects’ were exempt from competitive tendering.

We found two contrasting approaches to competition between NHS trusts and ISTCs. It was more common to make ISTCs subcontractors of an NHS acute trust (Shire, Metroland, Milltown), with Metroland PCT initially requiring two of its NHS hospitals to refer a percentage of patients onwards to a selected ISTC. Porttown, however, set up head-to-head competition between its NHS trust and an ISTC and, as a demonstration of intent to the local hospital trust, used its ISTC contract to the full. Other studies show that some165 but not all246 other PCTs also did this. The prospect of losing income to competitors made NHS trusts negotiate more constructively with their commissioners about service changes, for instance in audiology (Shire), ophthalmology and orthopaedics (Porttown).

Consultants’ conflict of interest over private work, reported nearly 20 years ago,250 was still evident. The Porttown ISTC employed its own consultants. Its manager claimed that, insofar as NHS consultants could, through their links with the referral management scheme, influence what choices GPs offered patients, they encouraged referrals to the other local private hospital, where they practised privately. The manager of the large private hospital in Metroland had

an agreement with the [NHS] trust that if they have a consultant who deliberately seems to be keeping their clinic to a minimum so that it will default to the private sector the trust will ring me so that the consultant doesn’t benefit from this behaviour.

Corporate hospital manager, Metroland

She ‘wanted to nip that problem in the bud’ in order to build a strong relationship with the PCT and not endanger her firm’s reputation for probity. Nevertheless, it was not in her hospital’s commercial interest to help the NHS reduce waiting times to the same as those for private hospitals, because that would be a disincentive for patients to buy health insurance and use private hospitals. Consultants using her hospital had the same incentive and would switch patients to other private hospitals if hers appeared to give NHS patients the same access as private patients. Tariff payments were lower than the premium prices once paid to private hospitals, which risked having to pay any costs above tariff level. Some Metroland GPs formed surgical provider groups, sometimes with consultants as members, and were commissioned by the PCT; a US study suggests that such hospitals stimulate demand for complex procedures. 251 Shire PCT found it harder to ensure uniform pathways across independent-sector providers than across NHS acute trusts. NHS trust clinicians were also under managerial pressure to meet NHS waiting-time targets; corporate providers were not. Private mental health services were mainly provided by the third sector, including a service user group (Metroland) and in Milltown a social enterprise involving leading commissioning GPs.

Attributing impacts to provider competition is not straightforward. After the ISTC opened, ALoS for orthopaedic patients in Porttown fell from eight days to four, the combined effect of competition and extra capacity. After hyper-acute stroke services were established in London, stroke mortality fell by 50% but in this case ‘provider substitution’ meant replacing existing treatments with evidence-based alternatives, not provider competition.

As case studies we selected PCTs both with and without multiple hospitals. Like others147,160 we initially equated multiple providers with ‘possible’ provider competition. Our case study findings undermine that assumption. Citing patient reluctance to travel152,252 and difficulty in influencing GP behaviour, the two case study commissioners (Metroland, Milltown) in large conurbations with many other hospitals nearby did not regard hospital competition as feasible or desirable, Milltown describing itself as ‘landlocked’. Metroland commissioners argued that patients prefer going to their local hospital or mental health service. Shire had three hospitals plus a teaching hospital just over its boundary, but still made the case-load allocations reported above. ‘Possible’ competition was not actual competition. Only Porttown stimulated head-to-head hospital competition, but having no other NHS hospital within or near its boundaries, competition there was between the NHS trust, a small private hospital and the ISTC. Even so, Porttown commissioners felt that they had limited competitive leverage over their local foundation trust:

Commissioning is about a power struggle between clout of PCTs and a hospital that will never close.

General practitioner, Porttown

Juridical controls

The standard DH contracts were used everywhere, but they accommodated local variation and their mandatory contents varied by care group. However, they specified only service volumes, not service design, case mix, clinical pathways or (until late in the study period) feedback of monitoring information. Mostly the same contract framework and terms were used for all providers, except that independent providers were always paid the normal HRG tariff (Shire). Additional local performance requirements reflected any local operating framework, health needs assessment (e.g. in Shire) and local agreements about care pathway changes (Metroland). DH control over contract formulation relaxed towards the end of the study period, putting PCTs in a stronger negotiating position by giving them more negotiating latitude about, say, how the 18-week waiting target would be met (Porttown). Some local authority contracts (e.g. in Shire) were for as long as 25 years, making them difficult to terminate early. At the opposite extreme, Porttown awarded annual contracts, creating uncertainty for providers. Community and mental health service contracts were less complete than hospital contracts. Provider non-compliance was initially dealt with by informal discussion. For instance, when a private diagnostics company in Porttown ignored written notice that it had breached its contractual obligation to obtain PCT approval before directly marketing its services to GPs, the next stage was to involve the PCT board, placing the matter in the public domain with possible adverse media coverage for the provider.

More complete contracts made service provision less flexible. Providers became reluctant to do work (e.g. diagnostic tests in Metroland, laparoscopic hernia repairs in Shire) off contract. CHS providers in Porttown complained that the PCT criticised them for not delivering activities that were not stated in the contract. Commissioners were reluctant to pay for off-contract treatments (even when a consultant had already offered to treat the patient), including novel procedures for which no tariff existed (both in Shire).

Informants readily cited local authority standing orders, EU regulations and DH Cooperation and Competition Panel requirements for competitive tendering. Many believed that the DH contestability framework generally permitted PCT collaboration with one provider.

Care Quality Commission (CQC) investigations could also trigger service reconfigurations independently of local commissioners. A CQC enquiry closed one ED and transferred maternity services to another hospital, irrespective of Metroland PCT’s hospital contracts. Adverse CQC reports compelled Porttown PCT to decommission an inpatient learning disability service. Similarly, Monitor imposed its own requirements on providers irrespective of their contracts with commissioners. In response to a financial crisis, Monitor replaced the entire management team in a Shire hospital, which the PCT took as opportunity to make a fresh start in working with its replacement.

In summary, the reconfiguration of commissioning structures (see Chapter 6) was resulting in the negotiated order between commissioners and providers evolving gradually from a predominantly manager-to-manager regime to one with a greater medical admixture. The discourse used in these negotiations correspondingly became somewhat more medicalised. Medical involvement appeared if anything to reduce the weight that commissioners gave to provider competition. The financial incentive regime changed little, being anyway partly beyond commissioners’ control, and juridical control over providers remained marginal.

The lead commissioner approach

In our study sites the nearest PCT or the one sending most patients to the jointly commissioned provider acted as lead commissioner (Shire, Porttown). This happened:

1. For very small care groups (e.g. certain ophthalmological treatments in Metroland, assisted conception in Shire).

2. In dealing with large teaching hospitals in conurbations. Thus Metroland was an associate commissioner for one of the largest London teaching hospitals (‘it’s a Goliath . . . we are small’) and lead commissioner for a smaller one.

3. For larger but evenly dispersed care groups. Thus Porttown PCT was lead commissioner for orthopaedic work for its ISTC on behalf also of three neighbouring PCTs. Before 2010, London PCTs with the same local acute or mental health provider also commissioned them jointly.

Lead commissioning required the associate commissioners to trust the lead commissioner to uphold their interests. Thus, commissioners sending just a few patients to a third-sector provider did not bother attending monitoring meetings because they trusted Porttown as lead commissioner to uphold their interests. Metroland subscribed to a distributed commissioning arrangement in which the lead role rotated between PCTs. PbR also promoted financial transparency, hence trust, between collaborating commissioners. Porttown and fellow commissioners began sharing information about service redesign. Metroland and its three cocommissioning PCTs produced a common commissioning strategy that led in turn to more standardised models of care (e.g. similar home treatment teams for people with dementia), standardised contracts and cost savings (e.g. through sharing older people’s home treatment teams across two PCTs). Shire resolved its disagreement with a neighbouring PCT about access criteria to assisted conception services.

However, commissioning was sometimes only partly standardised and pooled. Even in Metroland, each partner-commissioner also had separate meetings with the mental and community health service trusts about clinical change, information, finance and performance management. Initially each PCT negotiated service volume and costs independently of the other PCTs. Indeed, the mental health trust asked each PCT to keep the details secret lest other PCTs demand the same. After PCTs were clustered, they became more open with each other about such matters.

When different commissioners’ financial or service requirements were too diverse, lead commissioner arrangements broke down. Metroland PCT discovered that a lead commissioner without financial problems was likely to be more financially generous to providers than its less wealthy associate commissioners wanted. Metroland PCT eventually withdrew from jointly commissioning a cardiac service because it felt it could not influence, and hence could not gain much from, lead commissioner arrangements. Shire PCT disengaged from a lead commissioning arrangement that did not give it sufficient clarity about its financial commitment and risks. Discrepant fees for different PCTs’ patients led Porttown PCT to disaggregate the commissioning of the small but sole available provider of eating disorder services. Other disputes between lead and associate commissioners arose because, reflecting their populations’ different needs, they wanted different methods for approving changes to services and decisions about individual patients’ care pathways. Shire PCT withdrew from a dispersed commissioning arrangement because the lead commissioner had unilaterally decided not to stipulate distinct contract terms for Shire patients.

Detaching from a lead commissioning group at times enabled a commissioner to negotiate better terms. When Metroland began commissioning a large teaching hospital directly, it had to pay more attention to and take a stronger role in negotiating with it, obtaining more favourable terms for managing the financial risks of referrals of severely ill patients. Similarly, by negotiating independently, Shire made more flexible contracts with its providers.

While commissioning power was concentrated by having a single lead commissioner for a specialty or condition, the consequences of the lead commissioner model of distributed commissioning still depended on how convergent the different commissioners’ interests were. When they differed substantially, commissioners responded by reverting to separate commissioning. Otherwise, they responded with more standardised service specifications and contracts, gaining countervailing power in dealing with large, even merged, providers.

Joint local authority–NHS commissioning

In our study sites, joint NHS and local authority commissioning was ‘joined at the top’ (p. 21):257 more collaborative at higher (strategic, planning, governance) levels than at the operational level of dealing with service providers.

NHS and local authority commissioning strategies were often coordinated through reciprocal membership of each other’s working committees. Metroland had borough-wide commissioning groups with local authority members, primary care representatives and one patient representative. There was also a Health and Social Care Liaison group, but its only governance function was to manage Section 75 agreements, so the NHS bodies regarded it as a consultative mechanism. Porttown PCT included a local authority representative in its QIPP monitoring meetings. A psychiatrist from the mental health trust and the CCG’s mental health lead GP contributed to local authority commissioning of mental health services. Local authorities and NHS commissioners consulted quite widely about their joint commissioning objectives. Metroland public health leads made a ‘concordat’ and a Joint Strategic Needs Assessment, focusing on reducing health inequalities, which became the basis of a joint commissioning plan. In Shire a joint commissioning board agreed on collaborative work programmes, including hospital discharge management and preventing orthopaedics admissions, although that did not lead to any concrete commissioning decisions. Shire Health Scrutiny Committee still interpreted its role as approving significant variations in NHS services, so one of the Shire hospitals appealed against the commissioners’ proposals for competitively tendering community dermatology and endoscopy services. A local strategic partnership in Porttown aimed its public health priorities at the most deprived areas with poor health outcomes and at leisure, weight management and smoking cessation. The partners intended this work to inform joint commissioning plans. In practice, though, it had little influence on services to reduce childhood obesity or the licensing of fast food and alcohol outlets, although they did jointly commission stroke services.

At a strategic level, commissioning was more often a joint activity for community and mental health than other services. Metroland, Porttown and Shire were already jointly commissioning mental health services (since 1999 in Shire). Metroland and Shire both used Section 256 funds to help finance local authority re-ablement services and in Metroland falls prevention, re-ablement, home support, crisis intervention and home treatment services. The NHS and local authority jointly revised the corresponding care pathways and cofunded health crisis-intervention teams in Metroland, learning difficulties services and voluntary organisations’ memory adviser services in Shire. Shire social care, community hospitals, rehabilitation services and interface (GPSI) services were jointly commissioned, as was the mental health ‘floating’ support service in Porttown. Shire council also contributed to implementing clinical commissioning programmes for stroke, CHD and diabetes.

Being implemented through separate line-management structures, the operational work of provider negotiations, procurement and monitoring were more distinct, reflecting the differences between NHS and local authority regulatory and accountability regimes, financial and planning cycles, and whether local politicians were directly involved. NHS commissioners had larger budgets (for instance £25M vs. £2M for mental health in Shire), but local authorities had more experience of provider competition, commissioning private providers and involving service users in assessing tenders. The nearer to operational level one came, the greater these differences, hence the difficulty co-ordinating the different commissioners and through them the providers. PCT commissioning in Shire was organised by localities, which were coterminous with those of the local authority, but the intermediate care teams commissioned by either organisation were still resourced differently. Ill-defined boundaries between health and social care for, say, re-ablement (NHS) and rehabilitation (local authority) were reported to create commissioning overlaps and duplication. Thinking that trying to pool budgets would cause long delays in commissioning or changing services, local government and NHS commissioners in Shire chose instead to move towards colocation and then full integration of services. They were jointly developing specifications for integrated services, but had not yet commissioned any when our fieldwork finished. Although Porttown council and PCT agreed about mental health service development, their agreement was not translated into combined (mental health) service delivery teams or shared IT systems. In Metroland, both health and social care commissioners saw their role as keeping people out of hospital, but a lot of the real decision-making occurred at senior level in health and social care commissioners separately, ‘so that’s where things start to come apart’. Measuring health and other service outcomes was a new activity for many local authority staff, who discovered that many relevant outcomes (e.g. quality of life) are complex and hard to measure. Metroland borough therefore used falls outcome measures developed by the NHS commissioning support organisation, and some developed for the DH national outcomes framework, collecting its own data (e.g. on older people’s experiences of care homes) to populate them.

To mitigate these differences, NHS and local authority managers at both commissioner and provider levels set up informal, network-like working arrangements. Metroland managers set up an intermediate care partnership without seeking formal Section 75 status for it. Shire PCT and local authority informally agreed not to divert their budgets to other uses without consulting each other.

Nevertheless, points of friction between local authority and NHS managers remained, impeding joint commissioning.

1. Differences in priorities reflected the different accountabilities of NHS managers and local government managers, and, still more, councillors:Local authority manager, Metroland

   [T]he big difference is in culture between the NHS and its way of working and local council and its way of working . . . the members of the council are elected members and . . . the NHS is a very different way of working. And, I think there’s going to be, you know, lots of work, as we go forwards, around that culture because it’s just the culture changes will need to happen.

   Metroland council found shifting from health-care scrutiny to commissioning partnership a big cultural change. Porttown informants reported deep-rooted differences between elected councillors’, PCT managers’ and CCG members’ priorities, so that joint commissioning requirements could not be agreed, nor joint action taken. TCS and QIPP had made NHS–local authority relationships more ‘spiky’ (GP, Porttown). Different local authority and NHS managerial cultures were also reported in Shire. There especially, the local authority was service provider as well as commissioner, which our NHS informants saw as a conflict of interest.

2. Reconfiguration of NHS commissioners meant that they were no longer coterminous with local authorities. Metroland and Milltown CCGs were coterminous with a borough, but the central support organisation (for commissioners) for Metroland covered eight boroughs, which our informants feared might be hard to reconcile with the council’s wish to focus on local needs and services. When the shadow Shire CCG replaced locality-based PECs, it was coterminous with the county council instead of with several second-tier local authorities. Porttown PCT became part of a CCG locality coterminous with several councils, not just the city council.

3. One party to an agreement unilaterally changed commissioning arrangements and/or services. In Porttown the council established a single point of entry to mental health services, onto which the PCT could not easily map its service-commissioning methods. Metroland borough unilaterally cut £250,000 from mental health contract funding, leaving the PCT to pay for services that the PCT considered should be local authority funded. The PCT’s quick response to news of proposed local authority budget changes pre-empted similar unilateral reductions in Shire.

4. Clinicians were sceptical about whether or not local authorities would commission better services than themselves. For this reason, Porttown GPs were reluctant to publish a tender for mental health services, believing it would precipitate a city council ‘takeover’. The mental health trust was equally sceptical, but still thought it had a better working relationship with the council than with the CCG. In Shire the PCT, local authorities and clinicians had wide-ranging disagreements about strategy, priorities, service specifications and commissioning intentions for mental health services.

5. There were divergent interests and priorities for secondary care, for instance over priorities for disinvesting from inpatient services in Porttown.

The transferred public health activities and staff were the least connected. In Metroland, the role of the public health department was described as primarily strategic, delegating the implementation of its plans to others (disappointing the local authority’s initial hopes that public health staff transferred to it would have implementation experience or expertise), in contrast to Porttown, where the city council delegated the commissioning of drug services to its new public health department. In Shire, a still more silo-like relationship persisted, with public health apparently operating almost (but not completely) in isolation. During our fieldwork Health and Well-Being Boards were still being set up and did not ‘go live’ until late in the study period.

Distributed commissioning involved a negotiated order between commissioners, supplementing the other negotiated orders reported in Chapters 6 and 8. In theory, collaborating commissioners disposed of stronger financial incentives and greater ability to exploit provider competition, but they did not make obviously greater use of either than non-collaborating commissioners did. In joint local authority–NHS commissioning, differences in the discourses that the two kinds of commissioner used in dealing with providers were also apparent. Juridical controls remained marginal.

Client-based commissioning

We defined ‘client-based commissioning’ as specifying and paying for services on the basis of each episode of care for each individual patient (tariff payments, personal health budgets). Chapter 7 reports our main findings about tariff payments in England during 2010–12. As an initial exploration of whether or not service outcomes were associated with provider competition under client-based commissioning rather than population-based commissioning, we repeated the cross-sectional analysis, testing for associations between our 16 available policy outcome data sets and the three indicators showing the extent of provider competition in terms of client-based commissioning. Of the 48 predicted possible correlations only 1 was found (see Table 26, Appendix 4). The presence of a personal budget pilot scheme was associated with an increase in patients being treated within the 18-week target, but the association was weak. Otherwise the associations had the wrong sign, or were negligible or absent. Once again, the control variables were more often, and in this case also more strongly, associated with the service outcomes studied.

Negotiated agreement

NHS trusts’ responses to commissioners’ demands were never exclusively protectionist. They sometimes made alternative proposals for achieving results that the commissioners also wanted or, at least, were compatible with the commissioner’s priorities. Shire and Metroland hospitals were content to divert patients away provided the necessary alternative community and (social) care was available. Discussions at Metroland solved the problem of how to treat mental health patients who turned up at the ED in the absence of a 24/7 mental health service. Clearly specified, ‘micro-commissioned’ care pathways and common clinical interests were often enough to make providers respond constructively to commissioners’ proposals, especially if additional payment was also available.

Providers’ ability to respond

Even when provider managers agreed with a commissioner about what should be done, they might still be unable to implement their side of the bargain, for instance if consultants were unenthusiastic. In practice, pathway changes and bed reductions could not be made without clinician involvement. Shire hospital informants in particular explained that the hospitals’ ‘corporate’ view was that of the senior managers and consultants who dealt with the PCT, not necessarily that of other hospital staff. Repetitive discussions in Porttown care pathway groups led PCT representatives to doubt if the commissioners’ proposals were ‘getting back to’ hospital consultants (commissioning manager, Porttown). Consultants’ traditional autonomy from management worked against such involvement. In one Shire hospital especially, the consultants had customarily had a free rein to the extent of agreeing activity changes with the PCT without senior management approval. Milltown consultants took legal advice about who was responsible for the safety of patients triaged away from the ED, until their medical director persuaded them that this was a non-problem. In Porttown community services, consultants, psychologists and psychotherapists were reported to be very powerful, and managerial leadership weak, to the extent of its chief executive asking the PCT’s negotiator:

how he could bring about transformation in his organisation, which didn’t inspire hope in me I have to say.

Commissioning manager, Porttown

In a Metroland hospital, distrust between hospital managers and consultants had reached the point where a consultant believed:

Just about the time I started here, or just before, they shut the consultant common room because they were concerned that the consultants were using it to meet and plan the downfall of the executive, literally, at least that’s the story . . . you know, the agendas of consultants are not always the same as those of an executive board.

Surgeon, Metroland

Consultants were consulted about commissioning activity and ‘what we are required to deliver’ (hospital manager, Metroland), but not about curtailing or cutting services.

We found no evidence that NHS trust managers responded to other trusts, commissioners or private providers predominantly by developing competitive strategies, even when (as in Porttown and Shire) they knew that national policy favoured competition and there were ISTCs nearby. Rather, as indicated above, they tended towards more collusive, negotiative strategies (often with tacit commissioner endorsement) to circumvent competition.

Providers’ willingness to respond

A minor though common irritant to providers was commissioners’ demands for detailed information, which providers thought was clinically uninformative, ill chosen or not used once provided. When these requests became vexatious, one of the Shire hospitals merely ignored them. Provider-side informants sometimes interpreted the ideology of a ‘purchaser–provider split’ as endorsing their separation and independence from commissioners. Metroland trusts, we were told, had historically been quite inwardly focused and saw little need to engage with commissioner requirements. Personality clashes [such as the ‘massively adversarial’ relationship between commissioner and hospital chief executives in Porttown (GP, Porttown)] were another exacerbation.

When providers disagreed with commissioner suggestions for restructuring service provision, one response was to make incompatible counterproposals, as has been found elsewhere. 258 A Shire hospital proposed discarding services that it did not want to provide (surgery in community hospitals), protecting existing services (e.g. endoscopy) against provision by alternative providers (GPs in this case) and expanding other services. Especially when commissioner budgets were tight, some commissioner demands were simply not in the provider’s interest:

We call them efficiencies and to some extent they are but they are also service cuts, there’s no two ways about that, and it’s then you go from collaboration to confrontation.

Commissioning manager, Metroland

Shire commissioners were aware that they faced strong providers whose interests lay in maximising income and whose culture was that they, not commissioners, should determine service provision. Even after the medical director and consultants in one Shire hospital had agreed to reduce their high rates of cardiology intervention, hospital managers stopped them because of the financial implications. In Metroland, private care providers sought higher profit levels than the local authority was prepared to fund. Despite engaging in micro-commissioning, some providers (e.g. in Porttown) complained of being unable to influence commissioners. Indeed, the action research participants reported initial resistance by the providers to closer working between community and secondary care services.

When conflicts of interests became palpable, providers might, first, try to force the commissioners’ hand. As reported above, Metroland hospital reneged on its agreement to reduce its case load by 20%, knowing that PbR and AQP rules obliged commissioners to meet the costs anyway. This hospital also made unilateral changes to services (e.g. rapid-access nurse-led pain clinics, arrhythmia clinics) and then expected the commissioner’s post facto endorsement and funding. The main Porttown hospital unilaterally reconfigured stroke services to seek better integration with other providers. The smallest of three Shire hospitals threatened to withdraw its ED service, forcing the PCT to negotiate with another hospital to transfer the service there. Another hospital unilaterally informed the PCT that it would cease doing spinal treatments (it was the sole provider), forcing the PCT to retender the work. A third established a foot surgery department that increased its activity, even though another Shire hospital offered cheaper bio-mechanical alternatives. When Shire PCT could not afford to pay for additional patients to receive non-surgical treatment for prostate cancer, the provider obtained charitable funding and continued the service that way. Such tactics did not always succeed, however. A Metroland consultant raised money to build a cardiac catheter laboratory, but the commissioners refused to meet the running costs.

Arbitration between commissioners and providers rarely took place through any juridical process, although Milltown had agreed formal procedures for negotiating and resolving disputes. More often hospitals in Shire, Porttown and Metroland attempted to go over commissioners’ heads by asking the SHA to overturn agreements with the commissioner. The SHA could impose its preferred solution because it could sack both the provider and the commissioner chief executives if they did not implement it. Similar patterns are reported elsewhere. 39 At times SHA interventions weakened the commissioner’s bargaining position, for instance by imposing on the commissioner – but not providers – a deadline for agreeing contracts.

[Porttown] hospital has always won partly because the SHA undermined the PCT by covertly subsidising [the hospital]’s poor financial problem.

General practitioner, Porttown

However, another SHA refused to do that for the main Metroland hospital.

The commonest response on either side, though, was inertia: the often-reported phenomenon of seemingly endless, unproductive negotiations, even when much joint effort had been put into developing service plans, for example when it was proposed to cut services or change CQUIN payments (Metroland). As an extreme case, the only tangible outcomes of 10 years’ negotiation about ED services in Shire were a patient leaflet explaining the different purposes of EDs and GPs, and training care home staff not to telephone 999 but to use other providers. Especially if there were funding implications (e.g. when setting up an urgent care centre in Metroland), a negotiated agreement often had to be referred back to higher managers on either side, who might not endorse it. NHS providers were not necessarily enthusiastic about acting as autonomous public firms. 253 Corporate providers also had their moments of inertia (e.g. when Metroland PCT tried to renegotiate better terms and when Porttown tried to improve private hospital compliance with the 18-week wait target). Their culture of separation from NHS commissioners naturally ran deeper than in NHS trusts.

In a situation of mutual dependence, a commissioner with only one provider to hand might alternatively capitulate. Thus, Metroland PCT was forced to renegotiate its unrealisable admissions plans partway through the financial year. In the face of consultant opposition, Shire PCT turned down the GP initiatives mentioned above, which, a GP told us, was typical of the PCT’s tendency to placate and prioritise secondary care interests. Falling short of capitulation, Milltown commissioners developed a practice of trying to understand, rather than simply blame, providers whose actions or decisions caused problems for the commissioner. Indeed, a powerful, financially secure hospital trust could come to see its main commissioner as barely relevant to its own plans for strategic and service development.

Making providers more helpful

Providers’ responsiveness to commissioners thus reflected the presence or absence, and depth, of any conflicts of interests between the two. Providers often responded passively, but commissioners could sometimes rouse them into helpfulness by threatening loss of resources (possibly to a competitor) or a third party’s intervention. Porttown PCT had difficulty persuading a local hospital to provide wet age-related macular degeneration treatment as an outpatient service until it threatened to commission an independent provider. The same PCT discovered that, when the mental health trust felt that it might be about to lose resources or services, it became more willing to listen to the commissioner. In Milltown too, the threat of withdrawing the mental health trust’s psychology and rehabilitation service contracts triggered more constructive negotiations and micro-commissioning. More subtly, the Milltown commissioner hinted that, if the community health service specification were ‘refreshed’ in the ways the commissioner wanted, there would be no need to retender the service. In Shire audiology and mental health services, the threat of losing services (hence income) to other providers made the existing providers’ negotiators become more helpful. The availability of a competing provider was one way to make the threat of losing funds credible, but not the only way (see Chapter 7, section Provider competition). The absence of any existing vested interest also gave the commissioner a free hand, for instance when remodelling eating disorders and other non-psychotic mental health-care pathways in Porttown.

Soft coercion took the form of claiming that the SHA would intervene – to everyone’s detriment – if no agreement were reached: a credible threat, as explained above. Shire and Porttown PCTs both argued that the SHA would reduce budgets if QIPP targets were not reached. National financial targets were used as an argument to persuade the Porttown hospital to change its activities. In Milltown the lead GP commissioner represented herself as a ‘buffer’ between GPs and city-level commissioners, representing her own ideas as a lesser evil than what the city-level commissioners might otherwise do.

Our informants did not dwell upon the potential conflict of interest between GPs as commissioners and GPs as providers, although national media and the professional press raised this issue at intervals. Our GP informants who did mention it (e.g. in Metroland) perceived it in terms not of conflict of interest, but of opportunities for general practice entrepreneurship in developing primary care and substituting for secondary care, with little sense of any possible juridical or political implications.

Planned orthopaedics

For many of the more common orthopaedic procedures, measures and outcome indicators were relatively well defined and standardised, as were diagnostic criteria, care pathways, treatment methods and expected outcomes (although not, we were told, for every treatment, e.g. shoulder surgery). Our informants reported using PROMs data (mandatory), the National Hip Fracture Database (except Metroland), ALoS data (NHS Information Centre secondary uses service) and some generic measures applicable to orthopaedics (e.g. the Patient Environment Appraisal Tool) to compare providers’ activity, spending and performance and, in Shire, to assess the risks of reducing orthopaedic surgery provision in order to identify contract priorities. Metroland PCT also monitored hospital infection rates, but reportedly did not react to them. Shire PCT collected GPs’ reports of patients’ reactions to orthopaedic care, including patients who subsequently regretted having had ineffective surgery, with a view to introducing physiotherapy-based advisory services for patients contemplating orthopaedic surgery. For orthopaedics:

[Y]ou can absolutely quantify every stage of that, know what it’s going to be and for 96.4% of the time that will be exactly what happens and then the rest of it there’ll be a few complications.

Commissioning manager, Porttown

Because planned orthopaedic treatments were mostly rather ‘commodified’ it was relatively easy to establish criteria for appropriate referrals and to screen GP referrals accordingly, as Shire, Metroland and Porttown did, in effect making a single point of entry to planned orthopaedic services. Commodification also facilitated the substitution of one provider for another. Indeed, a corporate hospital in Metroland was for a time the PCT’s second largest orthopaedics provider. The predictable costs of much orthopaedic surgery, its high volume and a sufficiently generous tariff meant that:

elective orthopaedics is a money spinner.

Hospital manager, Shire

This applied both for NHS trusts’ ISTCs and for other for-profit providers over a long period. 250 Porttown fully used its orthopaedics contract with its ISTC to give the commissioner (and only incidentally GPs or patients) a choice of provider. Similarly, Shire PCT commissioned three corporate hospitals to do (mainly) joint replacement so that it could stop paying local NHS trusts extra each spring and summer to catch up with the orthopaedic work they had cancelled during the winter. In Metroland, though, both the private hospital and the ISTC were the subcontractors of local NHS trusts, not of the PCT or CCG (see Chapter 7, section Provider competition). However, corporate providers tended to treat only the less complex patients [American Society of Anesthetists classification 1 (ASA 1) to the stable end of ASA 3 comorbidity groups] in Porttown and Metroland. While some Shire NHS trust managers would also have preferred to concentrate on these patients, the consultants favoured a broader, more complex and varied case load.

Commissioners did renegotiate care pathway redesign for orthopaedics, but micro-commissioning had neither the centrality – compared with competition and financial incentives – nor the depth of work found with two other care groups (mental health and recurrent unplanned admissions). Commissioners thus controlled providers of ‘commodified’ planned orthopaedic surgery mainly by a combination of discursive control (evidence-based medicine) and provider competition.

Recurrent unplanned admissions

Commissioners’ efforts to reduce unplanned hospital admissions were impeded by a lack of knowledge of the categories and causes of these admissions. Shire and Porttown commissioners felt unable to predict even the direction of change, let alone number, of unplanned admissions in the coming year. Hospital informants in both sites stated that the underfunding of community services hindered them from discharging unscheduled patients, but PCT funds were mostly committed to hospital services. Shire GPs also perceived community services to be insufficient and so for safety referred older people who fell to an ED. In general, hospital services were commissioned per speciality or group of specialities; no one had dedicated overall commissioning responsibility for people prone to recurrent unplanned admissions.

Competition was of slight use to commissioners for influencing how providers managed this care group. If they required orthopaedic surgery, Metroland patients with long-term conditions and comorbidities were for safety unlikely to be referred to private providers, who merely took the less complex cases. The need to co-ordinate care pathways across multiple organisations was also, our informants thought, hard to reconcile with provider competition. Micro-commissioning was required. Thus the local authority and the mental health trust collaboratively commissioned re-ablement services (Porttown, Metroland, Shire), the Stroke Association to provide postdischarge care, home treatment for end-of-life care (Porttown), case management, carer’s services, intermediate care beds (Metroland) and GP triage of patients entering EDs (Shire, Porttown). In doing so, Shire reduced the number of re-ablement service providers from around 100 to 10. It also supplemented the contract of one local NHS hospital to pay for clearing the discharge backlog. However, both commissioner (in Metroland) and hospital informants (in Shire) claimed that they had difficulty dedicating sufficient time to the laborious tasks of redesigning care pathways for patients at risk of frequent unplanned readmission, especially if case management was envisaged (Shire). Each side recognised the value of such work and each argued that someone else should do it. In Livewell that someone else was a US HMO, engaged by the PCT (under pressure from general practices), which with those general practices comprehensively redesigned services for this care group, building in risk assessment and telephone support services. 238 Learning set members, however, emphasised that these integrative activities were not as routine as they ought to be, because of such impasses as the hospital asking the PCT to identify which patients to divert to community care, and to which services, but the PCT being unable to do so. Metroland PCT did not set targets for hospital discharge or providing services to prevent readmissions. Shire clinicians felt that the psychologists, CPNs and social workers were not much involved with the commissioners’ attempts to reduce unplanned admissions, so those attempts had little effect on working practice.

Prevention of coronary heart disease and diabetes

Coronary heart disease and diabetes prevention concerned two main types of activity and therefore provider. One was activity aimed at changing individuals’ consumption and behaviour patterns, addressing at-risk populations not in their capacity as patients (which most of them were not) but in terms of ‘lifestyle’, as consumers or residents. The other was clinical prevention, addressing at-risk individuals in their capacity as patients, whether undergoing routine health checks (e.g. under the national health checks programme), being treated for CHD and/or diabetes, or being treated for some other condition, and as necessary either treating the patient in general practice or referring the patient on to ‘lifestyle’ (e.g. smoking cessation) or other clinical services.

The commissioning of service providers was an activity ill adapted for stimulating primary prevention259 through ‘New Public Health’ methods – changing lifestyles and consumption patterns or regulating the production, promotion and sale of food etc. 260 – for CHD and diabetes as for many other preventable conditions.

[W]hen you do a needs assessment, this happens especially [with inter-]sector and other partners, they automatically think that you are going to commission a service and sometimes it’s not always about commissioning a service as such because if you’re looking at population health, you know, it might not be it.

Public health doctor, Metroland

One public health informant asserted that commissioning for ‘lifestyle prevention’ is simply not possible, whether through the Local Enhanced Services scheme or in any other way. Commissioners did not discuss long-term prevention much.

Because preventative services would not deliver immediate savings (e.g. within 3 years) they were not a high priority for commissioners, who thought that more pressing tasks such as tackling unplanned admissions had to be handled before moving ‘upstream’ to commission preventative activities. For lack of resources, few ‘lifestyle service’ providers were commissioned, although mechanisms such as the local enhanced services (LESs) provided ways to incentivise providers who did so. Indeed, the Porttown commissioner chief executive rejected public health proposals for the primary prevention of diabetes. Accordingly, public health staff there concentrated on trying to persuade the city council to be stricter in licensing fast food and alcohol retailers, and on carrying out health impact assessments for the council. Shire’s stroke and diabetes networks did succeed in commissioning exercise groups at sports centres, but the financial crisis led the PCT to withdraw finance for a diabetes-prevention pilot. These conditions might partly explain the reports that public health worked somewhat separately, a phenomenon reported beyond our study sites. 261

Secondary prevention, and prevention through clinical activity, was commissioned much as any other primary care clinical intervention, for instance through QOF incentives and LES incentive payments, as others have also reported. 259 Through its GP-owned social enterprise, Porttown PCT provided GPs with access to electrocardiograph machines and a toolkit to assist in vascular risk assessment, although the vascular checks petered out when the pilot project ceased. It also commissioned a small-scale project through which CHS nurses ran CHD and diabetes-prevention clinics in a deprived part of the city. Shire’s public health department set up a hub where GPs could send patients with mild diabetes or CHD for prescribed exercise or weight-loss services. A clinical commissioning project piloted the introduction of an intermediate diabetic care nurse and audited the clinical outcomes. The commissioners agreed to monitor hospitals on (among other measures) diabetics’ blood pressure and body mass index.

Mental health

Mental health service commissioners in our study all faced one near-monopoly NHS mental health trust absorbing above 60% (in Milltown, Metroland and Shire, 80%) of the mental health budgets alongside numerous niche providers, often local authority funded. (Metroland had at least 50, besides contracts for specialist tertiary services.) These small providers included voluntary and charitable providers, churches, counsellors and patient self-help groups, private care homes and domiciliary care. GPs and social services also provided substantial inputs (e.g. older people’s and learning difficulty care homes, day centres, domiciliary re-ablement).

Four media of commissioner power were noticeably weak in mental health care. One was discursive control. Mental health commissioners had only a small evidence base on which to draw. A few well-defined outcome indicators were available (e.g. return to activities of daily life), but our informants generally averred that outcomes were opaque and hard to measure, as well as difficult to relate to national policy, and monitoring data were usually incomplete. As for emic discourse, there were cultural differences between local authority and NHS models of the nature of mental health care and its intended outcomes, reflecting the different professional training and cultures of social workers and clinicians. Porttown, Metroland and Shire informants contrasted the recovery model (on the NHS side) that promotes patients’ active involvement in their treatment and tailoring services to the episodic nature of long-term mental problems versus a social model (on the social work side) offering indefinite low-key support (e.g. housing and employment) to maintain client independence. The medical director of psychiatry in Metroland complained that the social work ethos had gained ascendency, which ‘infuriates’ health workers and ‘disenfranchises’ clinicians. Second, the financial incentives were predominantly mainly block (on the health side) and per diem payments (for some local authority-funded services). The small-scale contracts stipulated patient volume, but seldom costings or performance measures. Against that, it was possible to discuss patient-level events in detail with small providers, which led to problem identification and solving. Monitoring of the large provider relied more on ‘generalisation and anecdote’ (commissioner, Porttown).

Mental health commissioners in all case study sites foresaw difficulties in using tariffs for mental health services, but also that tariffs would provide greater openness and clarity about provider activity and costs. The only scope for competition was among the small non-statutory providers, although Shire had opened negotiations with Mind and Rethink. Even among the smaller providers, though, Metroland PCT had never decommissioned an entire provider except for when a few temporary fixed-term contracts expired. PCTs’ limited tendering capacity and their aim to integrate health and social care were further disincentives to promoting competition.

Mental health care often required the assembly of co-ordinated care pathways across multiple providers, hence micro-commissioning and emic discursive control over providers. Micro-commissioning networks were relatively stable. Users’ voice was often more evident in these networks than in the commissioning of acute care, but users’ input did not necessarily translate into influence over budgets, the commissioners or the main provider. Mental health service commissioners in our study sites were more likely than their acute commissioner counterparts to include (besides managers) former service providers with social work, nursing or mental health provider backgrounds (Shire, Porttown, Metroland). Action learning participants, especially a GP commissioner, argued that too many psychiatrists spent their time medicating patients who do not get better. Consequently patients with drug and alcohol problems did not get access to psychological therapies.

The IAPT programme was an important exception to these patterns. IAPT contracts stipulated outcomes and limited access to services to a fixed number of contacts (20 in Porttown), giving a fixed end-point at which to evaluate treatment outcome. 262 Mental health commissioners in Shire and Porttown perceived in IAPT an evidence-based, outcome-oriented approach to commissioning services that might be adapted for dementia and personality disorder services too. Milltown commissioners also commented that it was

unprecedented . . . quite a sophisticated minimum data set.

Mental health commissioner, Milltown

The IAPT was thus a prototype for extending commodified commissioning further into mental health.

Technical contrasts

Table 8 summarises the empirical contrasts in NHS commissioning practice between these care groups.

These contrasts reflect technical differences across five dimensions between services for these care groups:

1. how far the commissioning of service providers is relevant at all (cf. primary prevention)

2. extent of the available evidence base

3. whether the episode of care and its outcomes are discrete and well defined

4. case mix homogeneity, hence, with (2) and (3), ‘commodification’

5. complexity of provision (number of providers simultaneously involved with one patient), hence the number of commissioners.

Provider responses to commissioners (cf. Chapter 10) also varied by care group. GPs responded to commissioners trying to strengthen clinical preventative services insofar as the services could be financed through LES, and above all the QOF, payment systems, but were otherwise not responsive. Few attempts were made to commission providers of intersectoral health promotion activity, so we had little opportunity to observe how they reacted. Providers of planned orthopaedic surgery responded more as the programme theory of NHS commissioning assumed, readily expanding case-loads when commissioners demanded, often competing with each other. In contrast, mental health providers tended to respond with renewed, or adapted, and collaborative micro-commissioning activity.

Germany

In Germany, the main health-care commissioners were SHIs (‘sick-funds’, Krankenkassen) and Land (provincial) governments. Eighty-seven per cent of the population were SHI members (2012), the remainder privately insured or self-payers. A separate SHI system finances long-term care. 265,266 Hospital ownership is diverse (public hospitals had 49% of beds in 2008, corporate hospitals 15% and charitable hospitals 36%). Acute care is paid for through DRG tariffs. In future, mental health care will be too. Land governments planned the allocation of hospital beds and largely financed the corresponding infrastructure. Patients can self-refer to any ‘ambulatory’ doctor, that is a generalist family doctor or non-hospital specialist. Ambulatory doctors were commissioned by dividing a cash-limited budget according to the points that each doctor earned, with different numbers of points for different medical acts. Reforms in 2004 required SHIs to promote a gatekeeping role for GPs to reduce direct access to specialist services, but in 2007 a survey (O’Shea L. The German Health System. London: Nuffield Trust [unpublished briefing paper]; 2010) showed that 44% of Germans saw two or more specialists in the last year (UK 19%). Cost control is a long-standing national policy. Table 9 summarises the mode of commissioning in the German system.

Managerial performance, negotiated order and juridical controls were the dominant media of power. 267

Italy

Factors influencing commissioning practice

Comparing the case studies of Germany and Italy with those in England, certain patterns recur, suggesting that these patterns reflect not local peculiarities of commissioning practice but common characteristics of the different media of governance across the three health systems and, on the basis of that evidence, corresponding lessons for NHS commissioning.

To overcome the information asymmetry between providers and commissioners requires both suitable resources (as the German case suggests) and (as the Lombardy case suggests) political will on the part of the commissioners and of the higher-level bodies to which commissioners are accountable. In all cases the negotiated order between commissioners and providers was coloured at local level by the micro-political relationships between organisations and among individuals, and by the history of past commissioning practice. Negotiations between commissioners and providers required a common discourse. Evidence about clinical practices and models (etic discourse) was one such discourse, although the extent of its use, which organisations most used it, and how, varied considerably across the three health systems. The content of shared emic discourses also varied considerably, reflecting the wider political cultures in which the three health systems were embedded: obedience to ‘policy’ dominated emic discourse in the English NHS; a solidaristic consensus on health policy and regulations generated by the ‘Rhineland’-style political institutions in Germany; and an ideology of subsidiarity in Lombardy. What the three situations had in common, though, was that the shared discourse enabled commissioners to appeal during negotiations to normative assumptions that the providers shared and that moderated the expression and pursuit of the particular interests of an organisation or profession.

In all three countries, DRG tariff payments gave providers strong incentives to gain income by expanding their activity. Whether or not they expanded competitively, at other providers’ expense, seemed to be a secondary question (especially in England and Italy; less so in Germany). This incentive, however, conflicted with cost-control pressures, already present in England and Italy, and emerging in Germany. Commissioners in all three systems – and Poland286 and the USA287 – reacted similarly by bundling tariff payments to weaken tariffs’ expansionary incentive effects on providers and to place an eventual cash limit on commissioner spending. Insofar as tariff payment systems are equated (simplistically, to be sure) with ‘competition’, commissioning practice in all three countries has involved sacrificing ‘competition’ to cost control. 276 In all three systems, a ‘pure’ tariff system tended to weaken commissioners’ control over providers, especially (but not only) by weakening commissioners’ power to select providers. Commissioners in all three health systems also supplemented tariff payments with separate incentives for maintaining clinical quality standards. To promote the ‘integration’ of care for complex and/or chronic conditions across multiple providers, commissioners in all three systems tried to introduce some form of micro-commissioning, although there was less scope for it in Germany and Italy than in England. In all three countries, public- and third-sector providers participated in micro-commissioning but corporate providers hardly at all in England (with the important exception of the US HMO in Livewell).

All three systems had organisationally diverse health-care providers. Irrespective of different modes of commissioning, state-owned tertiary providers tended to concentrate on complex and hard-to-treat (e.g. multiple) conditions. The third sector also provided especially for hard-to-treat conditions, but mostly in community settings (primary and social care). Corporate providers specialised in profitable care, typically high-volume non-urgent acute care, but also some more complex, high-cost acute care (e.g. heart surgery in Germany; complex diagnostics in Italy). However, because for-profit providers preferred lucrative, unproblematic services, they were less likely than public or third-sector providers to be providing services exposed to the causes of conflict noted above, despite their in-built motivation to expand their profits, hence market share, and despite their objectives differing most from commissioners’ objectives.

Commissioning practice always combined different media of commissioner power over providers. That reflects a certain hybridisation of the three systems studied, where Bismarckian elements (above all, tariff systems) coexisted with more Beveridge-like elements (attempts to maintain public accountability and implement regional or national health policy). Managerial performance and negotiated order were important everywhere, but thereafter the relative importance of ideology, financial incentives and competition varied. An elaborated negotiated order was an important, and for commissioners a helpful, part of commissioning practice, but it was not always easily reconciled with provider competition. Similarly, the transparency of providers’ activities and costs helped commissioners exercise control, but was not always reconcilable with provider profit seeking and competition. The dominant mode of commissioning reflected certain quasi-market structures: who owned the commissioners; what kinds of provider (in terms of ownership) were present; the rules for provider entry and exit; and how much discretion the legal and regulatory framework gave commissioners.

Summary empirical findings

Next we summarise our empirical findings, answering each research question in turn and stating the evidential basis of our answer. Briefly, three parallel strategies to governance coexisted in commissioning practice: a strategy based on trust and relationality; a more incisive, less trustful strategy (willingness to challenge, even replace, providers); and making ad hoc workarounds and modifications to quasi-market structures. After stating what limitations and qualifications apply to the findings, we draw some more enduring, general conclusions.

Limitations and qualifications

Specific caveats about the above findings are already stated in the methods and findings chapters. Here we add only more general limitations and qualifications to our conclusions. Beginning with theoretical limitations, other theories, such as sociolegal theory and institutional economics, might also be applied to and tested against our data. It may also still be early to expect a commissioning system undergoing reorganisation during the study period to realise commissioning outcomes first formulated in policy documents during 2010–12, although as Chapter 1 indicates these policies have substantial continuities with longer-established policies.

We did not take our cognitive frame analysis of the programme theory of NHS commissioning as far as a full discourse-historical analysis, or examine how the programme theory is embedded in wider genres and ideologies.

Our findings have, we repeat, a limited empirical basis: a systematic comparison of commissioners in three countries, a specific selection of England-wide data sets, a policy and discourse analysis, and four (for some issues, five) interview-centred case studies in England. We focused on health services, mainly secondary care at that, excluding intersectoral public health, research, training, capital allocation and, above all, long-term care, which in England, Germany and Italy is mostly financed separately. At the time of this study CCGs were still developing. Many details and practicalities of their eventual work remained unknown. The same applied to psychiatric commissioning in Germany. Our cross-sectional analysis was limited to published data, which made only two media of power susceptible to quantitative analysis and left it an open question if other measures of policy outcomes might show greater effects of competition, and if a longitudinal study would. We analysed the WCC competencies only in the aggregate, not how far each competence was separately associated with the health policy outcomes that we studied.

The cross-sectional analysis of managerial data used pre-CCG data for a period 2 years before the case study fieldwork, raising the question of what implications the differences in period for the cross-sectional and the case study data have for the present findings. Obviously no such implications arise from the free-standing conclusions of the cross-sectional study. On balance our findings suggest that commissioning practice in 2010–11 was not so different from 2008–9 that the cross-sectional and the case study data report substantially different commissioning regimes and therefore cannot be combined. Managerial staff turnover was high in our four case study sites but we found little evidence that commissioning practice changed substantially because of that. Indeed services there were reviewed and recommissioned less frequently. There was considerable continuity of GP involvement during 2008–11. If anything, GPs were in those sites less inclined than managers to use competition as a means of influencing providers, but, against that, the government became more inclined to after 2010. Formally, PCTs, not CCGs, were responsible for commissioning throughout. Nevertheless, the ideal solution to this time difference would be to extend both analyses into the period after 2010, yielding a longitudinal study, should the data become available.

Our mixed methods contributed unequally to the findings. The core research was case study findings at four English sites, with some in-depth work on exemplar tracer conditions and some comparative international case studies. Case studies are inherently open to recall and response bias, but triangulation between informants and other data sources reduced the risk. Comparing a variety of commissioners had the methodological advantage of enabling us to contrast varied organisational structures, but at the price of having to rely on the qualitative generalisation of our findings, a step to be taken cautiously. Our case studies covered four of the (then) 152 English PCTs, plus a fifth (Livewell) in part. Fewer case study data, in particular from providers, were available from one site (Milltown) than the other three. In Milltown the three main negotiative orders (GPs with commissioners; commissioner with hospitals; micro-commissioning) were long established and, despite other hospitals being available near its borders, the commissioners made little use of provider competition. If anything, the lower representation of Milltown might have led us to understate the importance of negotiated order versus provider competition as a medium of power. Another empirical limitation is that we did not observe formal commissioning meetings. However, this is less of a limitation than might appear. We found that formal commissioning meetings were not the only venue for commissioning work. Information and ‘intelligence’ sharing, not to mention ‘backroom’ dealing (p. 253),102 mostly occurred elsewhere. In one-to-one interviews, informants were freer to speak candidly and critically than in meetings. Discourses, ideologies and evidence are the same in meetings as elsewhere. In two sites, some of the pivotal meetings were so heated that it is doubtful outsider observers would have been admitted, and if they had been their presence would probably have had Hawthorne effects289 on how the meetings were conducted. From the case studies we have abstracted qualitative generalisations, which will apply more widely only to the extent that the commissioning methods we reported have counterparts elsewhere. However, because NHS commissioning is quite centrally controlled, pursues largely nationally defined objectives and has essentially similar organisational structures across England, our findings may be generalisable with caution. Table 15 shows that for the variables most relevant to commissioning practice and its service outcomes, our study sites (2010–12 data) covered most of the range of variation for England while avoiding outliers, except perhaps for PCT surplus/deficit, where our sites fell within a narrow range (small surpluses).

Most but not all of our study sites used private-sector commissioning support, also matching the England-wide pattern. 290

Nevertheless, comparisons with the rest of England, and of the UK, might yield different findings. At cross-country level, we compared just two other countries, although again selecting countries whose health systems have parallels elsewhere. Our selection of study sites meant that this study reported three main generic modes of commissioning, but others exist, including:

• general practice-based commissioning, under which a general practice, or other primary-care gatekeeper, holds the budget for their use of those services

• commissioning of health care for occupational groups

• corporate health insurers’ commissioning of health services

• commissioning of intersectoral health promotion (smoking-control campaigns, etc.).

Commissioning contingencies: environment, technology and ‘fit’

Empirical findings about CCGs and PCTs have a transient local interest, but taken with the cross-country comparisons they can make a more enduring contribution to explanations of which modes of commissioning and which commissioning practices emerge in different settings, and therefore how governance occurs through commissioning. Our empirical findings suggest a contingency account of how governance is exercised through different modes of commissioning. This study contributes to the study of governance by suggesting outlines of such a theory of commissioning, something which barely exists yet. Pending fuller discussion elsewhere, we note two main sets of external constraints on the modes of commissioning that a commissioning organisation can use: environmental constraints and technical constraints.

The media of power available to commissioners depended on the policy, legal and economic settings and the quasi-market structures within which commissioners worked i.e. the commissioners’ organisational environment. Combining our English (see Chapter 7), German and Italian evidence (see Chapter 12), we distinguished three generic modes of commissioning. Instances of all three were found in each health system, but one tended to predominate. Each mode of commissioning involved a specific configuration of media of power. Certain media of power, in specific forms, were available to commissioners while others were weak or absent (see Tables 9, 11, 13, 14 and 31). In the case of England, our four case studies and the cross-sectional analysis together suggested that only in quite specific circumstances did each medium of power appear to help commissioners exercise governance over providers. Governments’ willingness to face the electoral consequences of confronting providers (see Chapter 5, section Policy-makers’ and top managers’ programme theory) constrained commissioners’ negotiating position; German SHIs were constrained by providers’ legal rights to payment; and so on.

A second set of constraints arose from the technical character of health care for different care groups. Our evidence, mainly from five English case studies (including Livewell) and Germany, suggested distinct patterns of commissioning for different care groups. For each care group, commissioners had selected and developed configurations of the media of power which they considered to be adapted and practically relevant for maintaining commissioners’ governance over the corresponding health services. The technological character of different health services (e.g. for different care groups) had influenced the selection and impact of different media of power – not, as we had expected when writing the study protocol, technology at the level of differences in clinical or therapeutic practice, but at two other levels. One level was that of the characteristics of the model of care typical of that care group. Relevant characteristics were the extent of the available evidence base, case mix homogeneity, whether or not care outcomes were well defined, whether patients usually required a discrete episode of treatment or longer-term care involving multiple providers simultaneously, and whether or not health services or health promotion campaigns were relevant. The other level was that of the providers: minimum size compared with the scale of population health-care needs, which determines whether many providers were required, giving scope for provider competition, or just one; specialities; and geographical dispersal.

Contingency theory291 assumes that an organisation’s structures and managerial practices adapt to ‘fit’ its productive technologies and environment. Typically this adaptation occurs by trial and error. Observation of what adaptations develop, and under which circumstances, suggests which modes of healthcare commissioning appear adapted to which contexts and care groups.

We compared the media of power relevant to commissioning services for each tracer group (see Table 8) with those available to commissioners under each of the three modes of commissioning we studied (see Table 14). The comparison (Table 16) suggests that each mode of commissioning is a better ‘fit’ for some care groups than others. Some modes of commissioning make it easy, and others make it harder, for commissioners to exercise governance over the providers of services which use a particular kind of care ‘technology’. In Table 16, ‘Origin’ means the type of quasi-market setting292 which each mode of commissioning characteristically originated from and was adapted to. ‘Favoured providers’ indicates which type of provider organisations have tended to predominate and which, therefore, commissioners focus upon in that environment. ‘Care networks’ typically consist of public providers, small firms and third-sector organisations, who collectively provide and co-ordinate care for a group of patients. 293 Under ‘technology’, the ‘care groups’ row shows the care groups for whom the fit is closest between the technically relevant configuration of media of power and the mode of commissioning.

Episodic and chronic health problems are of course ends of a continuum. Developments in clinical and therapeutic technologies shift care groups along it, when for instance previously chronic conditions become more effectively treatable. The surrogate planning mode of commissioning appears applicable across most of this continuum for several reasons. Service planning at population level does not require the detailed patient-level service specifications or classifications that case mix commissioning and micro-commissioning do. Population-level norms of provision can be used instead, including norms based on epidemiological and other evidence that explains service utilisation patterns. The surrogate planning mode of commissioning can also accommodate flexible incentives (e.g. DRGs, target-based payments) at the margins of block contracts (though if the marginal incentives became predominant it collapses into case mix commissioning). Of the three modes of commissioning we examined in detail, only surrogate planning could readily accommodate the ‘technology’ of intersectoral health promotion campaigns outside the health sector (e.g. to promote exercise).

Emerging and adaptive commissioning: practical implications

A contingency theory of commissioning identifies factors, some beyond commissioners’ control, which constrain commissioners’ governance over providers. Yet any empirically credible theory must also accommodate our evidence (see Chapter 8) that commissioners nevertheless normally have some room for manoeuvre.

Commissioning practice was made to ‘fit’ the health system environment and/or health care technologies in various ways. Most obviously, commissioners selected or developed a mode of commissioning to fit the care group, their own aims and the health system context. Conversely, governments sometimes mandated a mode of commissioning in order to accommodate certain types of provider and/or care technology. However, care technologies, the demographic and epidemiological patterns which necessitate them, and the science which produces them are not very amenable to manipulation by policy-makers. Our findings suggest that, when there is no close ‘fit’ between organisational environment, care technology and their own aims, commissioners respond with one or more of the following:

1. Workarounds within an existing mode of commissioning. For instance, they tried to modify, amalgamate or invent DRGs for commissioning care for patients with multiple, complex long-term health-care needs.

2. Combining or hybridising several modes of commissioning (e.g. tariff payments with micro-commissioning in CCGs).

3. Supplementing commissioning with non-market governance structures (‘supplementary governance structures’ in Table 16). Thus, difficult contract negotiations in two of our English study sites became occasions when regional NHS bodies exercised hierarchical power (‘hidden hierarchy’). Commissioners in England, Germany and Italy tried to set up care networks to co-ordinate (‘integrate’) care across separate providers. Indeed, under micro-commissioning, contract-based commissioning was an adjunct to care networks rather than vice versa.

4. Neglecting certain care technologies (e.g. before 1990 many east European health systems had few means of commissioning community health services).

The emerging commissioning practices that we have reported were adaptive practical responses to the constraints facing commissioners. CCGs are also likely to have to develop adaptive responses insofar as commissioning policy rests on empirically questionable assumptions (see first section of Chapter 13). These emergent adaptations and their counterparts in Germany and Italy suggest some practical implications.

Our cross-sectional and English case study evidence suggested that provider competition had limited usefulness for governance. It was associated with commissioners achieving a minority, but only a minority, of the policy outcomes we studied, and it remained unclear what primary care providers had also contributed to these outcomes. Sometimes differently owned providers reacted differently to the same commissioning initiatives (see Chapter 7, section Provider competition). An obvious practical implication is that commissioner managers sensitise themselves to the differences in organisational goals, cultures and working practices between public, corporate and third-sector providers. For the ‘integration’ (co-ordination) of multiple providers, provider competition was of little practical relevance. As German SHIs had also concluded, micro-commissioning was also required, which depended on involving the relevant providers and then building long-term working relationships with and among them. Insofar as more diverse providers tended to have more divergent interests, it appeared a lesser evil to recognise the divergent interests rather than try to argue them away. As Williamson’s theory82 implies, these circumstances of asset specificity (of clinical skills), frequent contracts and complex, variegated products (uncertain outcomes) tended to impel commissioners and providers away from market-like relationships towards relationships more like those within a single organisation, although contracting mediated and masked them in quasi-markets.

In any case, when provider competition did occur (e.g. in Germany) providers’ resource dependency was what motivated it. Commissioners could harness that resource dependency directly, even when negotiating with a monopoly provider (see Chapter 10, section Making providers more helpful). Commissioners’ likelihood of success in doing so depended, though, on both sides believing that the commissioner could reduce the resources it allocated to non-compliant providers. However, the very fact that commissioners were exercising governance, that is acting as agents of the government’s, undermined this belief insofar as governments avoided publicly accepting responsibility for such acts (see Chapter 5, section Policy-makers’ and top managers’ programme theory) or when corporate providers were guaranteed NHS contracts. 246 In either event, commissioners ‘go naked into the conference chamber’294 to negotiate with providers.

It is often noted (see Chapter 2, section Negotiated order) that contractual relationships between commissioners and providers require a negotiated order – ‘relationality’ – alongside formal tendering, contract awarding and monitoring. We add that this is not only to compensate for contracts’ incompleteness. Neither is it a complete alternative to the more incisive commissioning strategies mentioned in the section Empirical patterns above. A negotiated order involving influential professionals on the provider side, evidence from our four English case studies (see Chapter 10) suggested, helps to promote understanding, hence willing implementation, of the commissioners’ aims in parts of the provider organisation which are not involved in formal contract or budget negotiations. Post-2010 NHS structures and rules also instituted broader, more systematic collaboration between consultants and GPs, placing GPs (as commissioners) on a more equal footing with consultants. This development, that of micro-commissioning and the strengthening of joint commissioning, attenuated the tripartite structural division which has inconvenienced interorganisational collaboration for the NHS since 1948. All these developments have the practical implication of commissioners adopting a long-term approach to maintaining trust between the parties involved, a task which involves both personal negotiating styles and the stability not just of the organisations but of the personnel involved.

We found in our four English case studies that a stable negotiated order between the ‘lead’ GPs involved in commissioning and the other GPs could be accumulated piecemeal by commissioners taking a flexible opportunistic approach to recruiting interested GPs when new policy initiatives or local project ideas arose. This necessarily happened gradually. Some GPs perceived that hospital interests had ‘captured’ local commissioners. Many GPs and consultants also wanted to avoid being made responsible for linking clinical and financial decisions, equated by some with rationing and commissioner scrutiny of clinical practice. An emergent solution was for commissioners to involve, if not ‘difficult’ GPs and consultants, then at least the stratum of (respectively) lead GPs and medical directors, in financial negotiations, for their support for commissioning decisions might be more persuasive to other doctors than non-medical managers’ opinions.

Besides the use of evidence basing, commissioner governance over providers through a negotiated order required a shared emic discourse. That the discourse is shared was almost as important as its particular content, provided that it legitimised commissioner governance over providers’ activity. Our English and German case studies suggested that norms which overemphasise providers’ independence were counterproductive for governance purposes. Insofar as commissioners can influence providers’ organisational cultures, the practical implications are obvious.

We found two emergent adaptations to client-based commissioning, more extensive in Germany than England or Italy. Our evidence suggested that unmodified tariff payment systems were likely to be counterproductive for the purpose of containing service costs. Tariffs were too marktkonform (market-conforming), that is too similar to prices in conventional markets, which tend to incentivise the open-ended expansion of production and allow only weak governance. 295 Commissioners in all three countries adapted the tariff system by defining ‘bundles’ of tariff payments and using non-tariff rates of payment at the margin, to incentivise a more controlled and focused development of case loads and case mix, and to protect providers that reduced case loads from penalty. Intelligently designed cost-and-volume contracts can be used similarly (Chapter 7, section Financial incentives). Non-tariff payments are anyway required for commissioning intersectoral health promotion. A second adaptation was for commissioners to try to influence providers indirectly by influencing patients’ choices through marketing methods (e.g. health advice services) or ‘nudges’ such as discounts for SHI subscribers, or by influencing GPs’ referral decisions. CCGs in our four study sites and Livewell (and elsewhere, judging by professional press reports) were increasingly attempting to manage GP referrals during 2010–12.

Monitoring information gathered in one commissioning cycle helps inform commissioners in planning the next. The new NHS structures and rules made hospital activity more transparent to commissioners and made commissioners’ plans for service development more explicit, specific and concrete. These are textbook precepts of good planning and project management. NHS commissioners nevertheless often depended on providers to collect and interpret monitoring data. A practical implication, especially observed in Germany, was for commissioners to develop provider activity databases whose data did not depend so heavily on the providers who were being monitored. This was a strength of DRG-based systems. If cost-and-volume contracts are used, the commissioners’ problem was then to obtain similar levels of transparency and information. An obvious practical implication would be to collate, preferably automatically, activity data that ought to be collected anyway at patient level for clinical management reasons, for instance referral data per provider by International Classification of Diseases code. Anonymised data usually suffice. German evidence suggests that it was effort well spent for commissioners to develop the specialised epidemiological and statistical skills for analysing the data. Yet our evidence, from both England and Italy, also confirmed the practical value of supplementing IT-based monitoring with qualitative knowledge of providers’ interests, problems and plans, through a negotiated order with providers. A striking example was the corporate hospital reporting consultant ‘misbehaviour’ (Chapter 7, section Provider competition). The acquisition of skills, links and relationships required stable staffing on both sides, an area in which clinicians’ knowledge and networks especially contributed to commissioning.

In summary, commissioning practice in our four case study sites worked in certain respects differently from what NHS commissioning policy assumed. Financial and ‘real-side’ contract negotiations were partly decoupled, with less clinician involvement on the financial side. Commissioners influenced providers (including fellow GPs) more through a negotiated order and discussions about evidence than through competitive mechanisms (including patient choice). Commissioners routinely compared providers against national and regional benchmarks, but seldom deselected providers for that (or any other) reason. Where multiple hospitals coexisted, analysis of cross-England data suggested that a minority of their service outcomes (including some proxy clinical outcomes) improved, but more did not. Different combinations of the media of power – distinct modes of commissioning – appeared to be required for exercising governance over the providers of services for different care groups, and for different organisational types and sizes of provider. Depending on these circumstances, governance through commissioning appeared to require giving different weights to three parallel strategies for exercising governance over health-care providers: one based on trust and relationality; a more sceptical, challenging strategy; and ad hoc workarounds when quasi-market structures seemed ill-adapted to commissioners’ purposes. A key managerial skill in the commissioners we studied appeared to be that of combining these elements judiciously. In any event, commissioners depended upon the transparency of provider activity data and the ability to analyse it. When commissioners and providers repeatedly have to work around quasi-market structures (e.g. by bundling tariffs, micro-commissioning), wider questions arise about how the health system should be structured.

Further research

Current English health policy assumes that private providers, social enterprises and NHS-owned providers respond differently from NHS commissioners in terms of ‘efficiency’, cost, propensity to innovate and willingness to compete. These differences are assumed to reflect the differences in provider ownership. We found some evidence consistent with this assumption and so, redactions notwithstanding, did the Competition Commission’s 2013 report. 296 It is important for the quality, development, integration and economical provision of NHS-funded care for NHS commissioners to understand what these differences in provider responses are; under what conditions and for which care groups these obtain; and if or how commissioners might deal with these consequences or indeed exploit them. Our evidence also showed the importance of micro-commissioning for certain care groups, raising the question of how to sustain effective micro-commissioning as providers increasingly compete and diversify. As noted, a limitation of the present study was that CCG commissioning had not yet ‘gone live’, making it too early to judge what effects CCG commissioning might have on the substitution of primary for secondary care, and upon cost control.

Our evidence, and its limitations, suggest the following needs, in descending order of importance, for further research to increase knowledge and understanding of:

1. if differently owned providers (corporate, third sector, public, professional partnership, etc.) respond differently from health-care commissioners and, if so, what specific implications for commissioning practice follow

2. the impact of commissioning practices on health system integration when care pathways have to be constructed across multiple providers that must tender competitively for work, perhaps against each other

3. the emergent system-wide effects accumulating, over a period of say 3–5 years, from the separate, competitive commissioning of an assemblage of providers of unequal size and diverse ownership

4. the negotiating practices used within contract negotiations, provider monitoring and problem-solving negotiations, not only in formal meetings but also in the informal micro-politics of deal making, tacit trade-offs and the sharing (or concealment) of activity data, cost data and ‘intelligence’ (qualitative, informal knowledge) between providers and commissioners. This is particularly relevant to monopsony–monopoly negotiations (i.e. a commissioner with its dominant provider).

Additionally:

1. What variants of ways of bundling tariff payments can be found and

   1. how do they compare as means of cost control?

   2. to what extent do they retain the monitoring-information advantages of classic tariff systems?

2. Data availability limited our reanalysis of managerial data to a cross-sectional study design. There remains a need for a longitudinal analysis, as data become available. Our evidence about the difficulty of interpreting such findings suggests that such a study ought to include qualitative research into the organisational mechanisms producing any important patterns found in the quantitative longitudinal study.

3. What commissioning practices and other conditions help CCGs and local government to focus and collaborate – despite our evidence of the difficulties – on managing certain strategically important activities (avoidable admission prevention; making planned care more efficient; primary prevention through intersectoral activity)?

4. Once CCGs have stabilised and matured, what continuities and discontinuities from PCT commissioning practice will they show?

5. What specific managerial skills, and other personal attributes, commissioning managers might require; for instance, which of the WCC competencies remain relevant to NHS commissioning.

Contributions of authors

Rod Sheaff (Professor, Health Services Research) designed the research, conducted the discourse and cross-sectional analyses, contributed to fieldwork in England, Germany and Italy, synthesised the results and prepared them for publication.

Nigel Charles (Research Associate, Health Services Research) undertook the project management and contributed to the English fieldwork, facilitating the action learning set, discourse analysis, English and German fieldwork, case study analysis and preparing the results for publication.

Ann Mahon (Senior Lecturer, Health Policy and Management) led the action learning set and contributed to research design, discourse analysis, English fieldwork, case study analysis and preparing the results for publication.

Naomi Chambers (Professor, Healthcare Management) contributed to the research design, facilitating the action learning set, discourse analysis, English and German fieldwork, case study analysis and preparing the results for publication.

Verdiana Morando (Research Fellow, Health Services Research) undertook the Lombardy fieldwork, analysed the Italian data, and contributed to the action learning set and preparing the results for publication.

Mark Exworthy (Professor, Health Policy and Management) contributed to the research design, discourse analysis, cross-sectional analysis, English fieldwork, case study analysis and preparing the results for publication.

Richard Byng (Professor, Primary Care Mental Health) contributed to the research design, discourse analysis, cross-sectional analysis, case study analysis and preparing the results for publication.

Russell Mannion (Professor, Health Systems) contributed to the research design, discourse analysis and preparing the results for publication.

Sue Llewellyn (Professor, Accountability and Management Control) contributed to the research design, discourse analysis, Italian fieldwork, case study analysis and preparing the results for publication.

Publications

Chambers N, Sheaff R, Mahon A, Byng R, Mannion R, Charles N, et al. The practice of commissioning healthcare from a private provider: learning from an in-depth case study. BMC Health Serv Res 2013;13(Suppl. 1):S4. doi:10.1186/1472-6963-13-S1-S4.

Sheaff R, Chambers N, Charles N, Exworthy M, Mahon A, Byng R, et al. How managed a market? Modes of commissioning in England and Germany. BMC Health Serv Res 2013;13(Suppl. 1):S8. doi:10.1186/1472-6963-13-S1-S8.

Sheaff R. Plural provision of primary medical care in England, 2002–2012. J Health Serv Res Policy 2013;18(Suppl. 2):20–8.

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. 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.

Published and grey documents analysed (excluding transcripts)

Equity and Excellence 2010. 195

Liberating the NHS – Allied Health Professionals. 301

Liberating the NHS – Analytical Framework. 302

Liberating the NHS – Commissioning for Patients consultation paper. 303

Liberating the NHS – Equality Impact. 304

Liberating the NHS – letter from Sir David Nicolson. 305

Liberating the NHS – Local Democratic Legitimacy in Health Consultation Paper. 306

Liberating the NHS – Regulating Healthcare Providers Consultation Paper. 307

Liberating the NHS – Report of the Arm’s Length Bodies. 308

Liberating the NHS – Transition – Nicholson annex. 309

Liberating the NHS – Transition – Nicholson letter. 310

Liberating the NHS – Transparency in Outcomes: A Framework for the NHS Consultation Paper. 311

Liberating the NHS – An Information Revolution Consultation. 312

Health and Social Care Act 2012. 196

A1. Factsheet – Overview – 24 April 2012. 209

A2. Factsheet – Case for Change – 24 April 2012. 210

A3. Factsheet – Overview of Health and Care Structures – 24 April 2012. 211

A4. Factsheet – Scrutiny and Improvements – 30 May 2012. 212

B1. Factsheet – Clinically Led Commissioning – 24 April 2012. 213

B2. Factsheet – Provider Regulation to Support Innovative and Efficient Services – 24 April 2012. 214

B3. Factsheet – Greater Voice for Patients – 30 May 2012. 215

B4. Factsheet – New Focus for Public Health – 25 April 2012. 216

B5. Factsheet – Greater Accountabilty Locally and Nationally – 25 April 2012. 217

B6. Factsheet – Streamlined Arm’s Length Bodies – 25 April 2012. 218

B7. Factsheet – On Support Worker Regulation – 27 February 2012. 219

C1. Factsheet – Improving Quality of Care [undated]. 220

C2. Factsheet – Tackling Inequalities in Healthcare – 27 April 2012. 221

C3. Factsheet – Promoting Better Integration of Health and Care Services – 27 April 2012. 222

C4. Factsheet – Choice and Competition – 27 April 2012. 223

C5. The Role of the SofS – 30 May 2012. 224

C6. Factsheet – Reconfiguration of Services – 27 April 2012. 225

C7. Establishing New National Bodies – 27 April 2012. 226

C8. Research – 27 April 2012. 227

C9. Education and Training – 27 April 2012. 228

House of Commons Health Committee. 2010. Commissioning. Fourth Report of Session 2009–10. London: HMSO. 30

Leximancer data tables, documentary and oral material samples combined

Leximancer analysis

Leximancer software, which automates quantified content analysis, proceeds unless otherwise reconfigured as follows.

1. The software divides the study text into two-sentence blocks.

2. It eliminates stop-words (proper names, ‘and’, ‘the’, interviewer name and other terms that are known a priori to be uninformative).

3. Frequent words and frequently associated words are selected as ‘seed words’ (‘concepts’).

4. The software codes the two-sentence blocks according to what concepts are present.

5. It counts the occurrences of codes.

6. The most frequently associated concepts are defined as themes (which can be traced back to their textual sources).

7. By default, Leximancer surfaces the main concepts inductively. The researcher can also rerun Leximancer further times, selecting those inductively found concepts that are relevant to her research questions for grouping into themes. In the present case, we selected the concepts and themes relevant to the power mechanisms discussed in Chapter 2.

The term ‘concept’ has in Leximancer analysis a narrower than usual meaning, denoting ‘co-occurring words’ (as opposed to the more usual use, in research, of a theoretically informed essential definition).

The study ‘text’ can be a combination of documents (including transcripts, laws), spreadsheets, audio and video material. We therefore ran three analyses:

1. Equity and Excellence: Liberating the NHS195 and the official support documents branded as explaining and elaborating it, except for the Assignment for Transition and human resources management documents,297,298 which were irrelevant to a CMO analysis

2. oral material: speeches and interviews

3. the 2012 Act196 with the official explanatory ‘factsheets’, including one on service quality. 209–228

First, we used Leximancer’s default setting to find inductively what themes (and component concepts) were present in the sample. We coded the concepts found as relating to the context, mechanism or outcome of NHS commissioning policy or as ‘stop-words’ (e.g. ‘change’, ‘future’, ‘things’, ‘and’, ‘the’, etc.) for being uninformative, ambiguous (e.g. ‘substitute’), trivial or irrelevant. Approbations, however vague (e.g. ‘best’, ‘improve’), were also coded as outcomes, that is policy or service outcomes. We collapsed duplicate concepts (e.g. ‘patient’ + ‘patients’, ‘GP’ + ‘GPs’, ‘better’ + ‘best’ + ‘improve’). We assumed that the conjunction of concepts or themes denoting a mechanism and/or a context and/or an outcome denoted an existing or proposed CMO relationship. A count of these conjunctions showed which CMO relationships received most coverage in the texts. Because some of these conjunctions (textual proximities) may reflect nothing more than drafting accidents, this method may bias towards overestimating the number of CMO relationships stated in the texts, but if any such overcounting is more or less evenly distributed across the texts, as we have assumed, it will not bias the relative frequencies of the different CMO assertions. From the blocks of texts where Leximancer had found these conjunctions, we extracted any descriptions of CMO relationships and classified the mechanisms according to which media of power they used.

Cognitive frame analysis

Although the quantitative (Leximancer) analysis located which CMO relationships the texts most often mentioned, they were too broad, ambiguous or brief in saying how these mechanisms worked or would work. We therefore made a cognitive frame analysis of data from our interviews with parliamentarians and top-level health managers to elaborate and supplement the accounts of CMO relationships found in the policy texts. In doing so, we again sought to relate the accounts and explanations (frames) that our informants used to the categories (CMO; media of power) required for the present study. Mostly the informants’ accounts were consistent, but where they differed (in emphasis rather than contradicting each other) we took the more often expressed view as the one more likely to guide commissioning in practice.

When policies are controversial, a simple précis of policy documents and transcripts is likely to oversimplify the programme theory by omitting relevant aims, mechanisms and implicit background assumptions. It might also overemphasise spurious rationalisations and the polemics. A more sophisticated discourse analysis is required. We adopted a ‘rhetorical’ variant. 293 This is ‘critical’ in the sense of not necessarily taking all managerial and political rhetoric at face value or as entirely coherent, valid and normatively persuasive. We therefore dispute the suggestion that such critiques are impossible;294 its realist character is just what gives realistic evaluation its critical facet, because a programme theory can be evaluated empirically. Therefore, to the extent that they rely on these empirical assumptions, so can the policies that through the medium of political discourse express a programme theory. Nevertheless, taking policy statements at face value is a necessary starting point and our default assumption until we find reasons to suspend it.

We took the following signs as calling into question if policy documents should be taken only at face value:

1. silences or obviously ambiguous policy positions about important mechanisms or outcomes

2. statements contradicting the balance of evidence available when the policy was formulated

3. apparent contradictions among policy statements that (all supporting the policy) ought to be consistent.

Then a realistic evaluator has to infer and impute the missing assumptions in order to reconstitute the programme theory as completely and explicitly as possible and (to avoid evaluating a ‘straw man’ theory later on) in the most credible form consistent with the explicitly stated elements. We did so by inviting policy-makers themselves to elaborate the missing material at interview.

We collated the descriptions of CMO relationships found by these methods and paraphrased them as statements of the form ‘Doing X in circumstances M will cause agent A to do Y’ (or a logically equivalent statement, e.g. ‘If A does X, B will do Y’), the form required for empirically testing CMO assumptions. In this way we identified the CMO relationships by which policy-makers and top managers assumed NHS commissioning would achieve its intended service outcomes.

Cross-sectional analysis of published managerial data

Before making the regression analyses, we checked for multicolinearity by measuring VIF among the potential independent and control variables in the regression analyses, retaining only variables whose VIF was below the conservative threshold of VIF = 2.5 (hence also the conventional threshold VIF = 5.0).

All analyses were at PCT level. To show effects of, say, provider competition, it is necessary (but not sufficient) to find non-trivial correlations, with coefficients of the correct sign, between the commissioner characteristics and the service outcome variables. Since there are a number of such variables, there would be multiple potential correlations for each commissioner characteristic for which we had data. The higher the proportion of such correlations found having the sign that the relevant element of programme theory predicts, and the stronger those correlations, the stronger would be the evidence supporting the assumption that provider competition has an impact on the policy-relevant service outcomes. Such findings would support the inference that, if PCTs could stimulate (continuing the example) provider competition, the PCTs would thereby help to realise those outcomes. Conversely, the absence of any such correlation, or the presence of correlations with the opposite sign from what the programme theories outlined above assume, would be prima facie evidence against those assumptions.

In the event (see Chapter 7, subsection Provider competition) we found few of the correlations that the programme theory assumed, in particular regarding competition. We therefore tested the robustness and sensitivity of our findings by rerunning the analyses for only the PCTs with the highest levels of competition, that is those in the top quartile for:

1. spend on independent (i.e. for-profit) sector

2. spend on local government sector and voluntary sector combined (not separated in the published data)

3. Herfindahl index.

The top quartile was selected because it contained 38 sites; a smaller selection would allow test results only of dubious validity.

Evidence synthesis

Across the case studies, data were synthesised by framework analysis. Conceptually this was equivalent to constructing, for each research question, a data grid in which each row contained data about a specific aspect of that research question, and each column represented a site, and then populating the cells with the relevant data from the case study collections of ‘pithy sentences’, findings from the cross-sectional analysis, action learning set findings, international comparisons and other published studies. We noted what common or divergent patterns there were across cells and then ‘read off’ the patterns as answers to our research questions. This method also revealed where it was necessary to add new categories or concepts to accommodate unforeseen empirical findings. By combining primary and secondary sources, we were able to compare (indeed check) our own findings against those from other studies.

As necessary, we derived the (equivalent to) row headings for each such systematic comparison from the analytical framework in Chapter 2, the programme theory assumptions found by discourse analysis (see Chapter 5) and by deduction from the research question itself. In this way we nuanced the framework analysis for each research question. Analysing the reconfiguration of commissioning structures (RQ2) required a comparison of longitudinal accounts of the formation and development of commissioning structures in each study site during the study period. Once these histories had been elicited, they too could be systematically compared in the above way. Regarding RQ3(a), the ways in which commissioners changed their commissioning practice in an attempt to influence their providers demonstrated that the commissioners had at least that much freedom of manoeuvre in practice. The limits to this freedom were found by discovering what practical, resource and policy restrictions there were on their freedom to exercise the media of power listed in Chapter 2 over their providers. We identified these limits from our case study materials, policy and regulatory statements. To examine some of the effects of client-based commissioning [RQ3(b)], we relied more on the cross-sectional than the case study data. Analysing the similarities and differences in commissioning practice for different care groups provided the basis for testing some of the theories discussed in Chapter 2 (RQ4). 204

Metroland

Metroland is a borough in the London NHS region with a population of about 270,000. It had been served by a PCT whose boundaries corresponded to those of the borough. Before fieldwork started, the PCT had clustered with three neighbouring PCTs, each of which was part of the same conurbation. During fieldwork this cluster had merged with a similarly sized neighbouring cluster.

The population served by the former PCT is in the second quartile of measures of deprivation. The main Public Health England average measures of health in Metroland are mixed compared with the English average. Life expectancy and healthy eating are higher than the English average, but smoking-related deaths and levels of children in poverty are higher than average. Premature mortality caused by CHD is the same as the English average.

There are about 50 GP practices in the borough. The former PCT largely dealt with two acute providers. During the course of fieldwork, one of these providers merged with two other acute providers in other parts of the conurbation. Given the PCT’s proximity to many other acute providers in the city, it also dealt with some of these. About 55% of its allocation was spent on one of the main acute providers and around 33% was spent on two others. Acute services were also offered by two private providers. The original four clustered PCTs shared services with one NHS community and mental health service provider. Some mental health services were also provided by the voluntary sector. The local authority boundaries corresponded to those of the original PCT, which also commissioned adult social and mental health-care services, some jointly with the clustered PCTs.

Subsequent to the clustering of the PCTs, demerging has in effect taken place with the creation of a CCG whose boundaries correspond to that of the original PCT.

Porttown

Porttown is a city with a population of about 260,000. The population served by the former PCT is in the top quartile of measures of deprivation, as measured by income and employment scores. 299 The city has above average poor health for most of the main measures published by Public Health England. For example, life expectancy in the city is lower than the English average and it also ranks below the average for measures of healthy eating and above the average for smoking and premature mortality caused by CHD. The proportion of children living in poverty is about the same as the English average. 300 The Network of Public Health Observatories produces health profiles for each local authority in the UK. These use Office for National Statistics data from the Index of Multiple Deprivation for 2010 for each Lower Super Output Area to describe the sociodemographic characteristics of these authorities (www.apho.org.uk/default.aspx?QN=P_HEALTH_PROFILES).

There are around 40 GP practices in the city. Secondary acute care is provided by one large district general hospital (DGH) and two private hospitals. The DGH accounted for over 90% of the PCT’s spend on its main acute provider. At the time of fieldwork, the PCT retained its public health provider arm, but the commissioning function was in the process of being transferred to the local authority. There is one main NHS provider of community and mental health services, including inpatient provision. There are also out-of-area inpatient services commissioned from the private sector. Mental health services are also provided by a small number of voluntary-sector organisations, some commissioned jointly with the local authority. The PCT boundary was coterminous with the city local authority. The local authority is a commissioner of adult social and mental health-care services.

As with all our study sites, the clinical commissioning group was in the process of being established during the course of fieldwork and in April 2013 it assumed commissioning responsibilities from the former PCT and that of a neighbouring one. The resulting CCG boundaries covered the city and most of the county in which the PCT was located, including another large city. The CCG is divided into three localities. One of these localities covered the former PCT’s catchment area, but also extended into the surrounding rural hinterland to serve a population of about 350,000.

Shire

Shire is a predominantly rural county with a population of about 400,000. Two former PCTs served this population, and had clustered before fieldwork. One covered a largely rural population and the other the main urban centre of the county, consisting of two neighbouring large towns with a combined population of over 300,000. The population served by the former PCTs was in the third quartile of measures of deprivation. The county population has generally better health than the English average. Life expectancy and healthy eating are above the average for England. It also has lower than average rates of smoking, premature mortality caused by CHD and children living in poverty.

There are over 100 GP practices in the county, with nearly half of these in the two main urban centres. There are three DGHs in the county, two of which were in the process of merging during fieldwork. The preclustered PCT that served the urban centre spent around 40% of its allocation on one of the main acute providers and around 90% on all three of the providers in the county. The PCT that had covered the mainly rural areas of the county spent around 36% and 70% respectively on these acute providers. Acute services were also offered by four private providers. There was one provider of community and mental health NHS services in the county, with some provision by the voluntary sector, some of which was jointly commissioned with local authorities. We were unable to gain access to the public health service provider. There were three local authority commissioners of adult social care and mental health-care services. The rural PCT boundaries broadly corresponded to those of the county local authority, and the urban PCT boundaries broadly corresponded to those of the local authorities for the two neighbouring towns.

The clinical commissioning group that was being formed during fieldwork now spans the boundaries of the clustered PCTs; that is, it covers the entirety of the rural and urban areas of the county. The PCT that had covered the mainly rural areas of the county had created a locality structure. This has been retained and adapted to cover the CCG, which is subdivided into 13 localities, each with their own commissioning responsibilities.

Milltown

Milltown has a population of 229,000 and is a city that is part of a large conurbation.

The population within the boundaries of the former PCT is in the top quartile of measures for deprivation. The health of the population is worse than the English average on nearly all of Public Health England’s main measures. Life expectancy in the city is lower than the English average and the proportion of children living in poverty is above the average. It has especially higher than average rates of adult smoking, smoking-related deaths and premature mortality caused by CHD and cancer.

There are just over 50 GP practices in Milltown. Most services are commissioned from two main providers: the acute provider and the mental health provider. About 68% of the PCT’s allocation was spent on the main acute provider and 19% on two others in neighbouring cities. Other services are commissioned from a range of voluntary, community and private-sector providers. Public health commissioning has been transferred to the local authority, which is coterminous with the former PCT patch.

The CCG had been working in shadow form for some time in advance of its legal status. It has assumed the commissioning responsibilities previously held by the Milltown PCT.

Livewell

Livewell was selected as a stand-alone single in-depth case study of a public–private partnership because of government policy to encourage private-sector involvement in commissioning. In this context it involves partnering with a private-sector organisation, as a cocommissioner and provider, to aid whole-system redesign in primary care and to support the commissioning of services for people with long-term conditions at risk of unplanned hospital admissions.

Livewell is an area of the West Midlands characterised by high levels of deprivation, communities with a high proportion of people from diverse minority ethnic backgrounds, poor levels of general health and high levels of chronic disease prevalence. In common with other areas, the health economy is challenged by high and increasing levels of ED attendances and urgent care admissions. Thirty-two different first languages are spoken in the community.

Across the five practices in the Livewell programme, 32% of urgently admitted patients stay in hospital for less than 24 hours; 60% of ED attendances result in patients being discharged, of whom 40% go home without treatment; and there is a large degree of variation in urgent admissions, ED attendances and outpatient referrals between the five practices. In one of the practices, where a more in-depth analysis has been undertaken, 54% of spend has been estimated to be on urgent care, in contrast with inpatient elective (16%), day cases (20%) and maternity (10%).

Driven by the doctors, there was also a series of events held with staff to shape the programme strategy and redesign. Patient participation groups were actively involved on an ongoing basis in design and in offering detailed feedback. Users and care-givers took up structured opportunities for engagement.

Phase 1 of the programme started in 2008 and covered one practice, of 9500 patients. Programme management support was secured from a private-sector health-care commissioning support provider, which also delivered the telephone care-management service. Phase 2 of the initiative, which was conceived in 2010 and sponsored by the local PCT as a fund-to-save initiative (approximately £500,000), comprised five practices covering 50,000 patients and was supported by the same private-sector firm. Redesign of care included an initiative to run group consultations, a think-tank and a telephone care-management service. It is this last, in phase 2 of the project, that is the principal focus of this case study.

Sharing preliminary findings and enhancing the research process

When the first action learning set met (2010), analysis of the first phase of case study site interviews was largely complete. From the analysis we constructed scenarios (Table 30) instantiating answers to the question ‘How does power manifest itself in commissioner–provider relationships?’ and presented them to the learning set. The ensuing facilitated discussion endorsed the face and content validity of the scenarios as instances of uses of the media of power. The scenarios served participants as a framework with which to explore notions of power and how they are enacted in commissioning practice. Participants found scenarios from sites other than their own both familiar and comforting, in that they recognised they were experiencing comparable instances of different media of power.

Further data analysis had taken place by action learning set 4, to which we fed back the emerging findings. The fifth meeting made a critique of a Nuffield Trust research report on commissioning. The learning set:

• endorsed the face and content validity of our research findings so far

• suggested adding ‘class hegemony’ and bullying into the media-of-power schema

• considered that the research provided a common framework for learning set members to explore their experiences, an opportunity for a more sophisticated discussion of experiences, problems and possible ways forward in the workplace.

Sharing commissioning experiences: international perspectives and those from other sectors

At the ‘mini-conference’, contributors from Techniker Krankenkasse (Germany) and Emilia Romagna health region (Italy) described their own commissioning practice and its impact on health care, offering ‘expert outsider’ insights into the focus, processes and expected effects of their commissioning efforts. We presented more of our research findings and there was a facilitated discussion about the applicability to health-care commissioning of supply-chain management models derived from manufacturing procurement.

The learning set concluded that health-care systems can learn from exploring commissioning and procurement practices in countries with different approaches to funding, structuring and delivering care. Health-care systems can also apply – with adaptations – private-sector procurement processes to state-funded and insurance-based health-care systems. The Cousins Supply Wheel62 depicts the different pressures related to effective supply-chain management (organisational structure, portfolio of relationships, cost–benefit analysis, skills and competencies and performance measures) in other economic sectors. The learning set members found it relevant to NHS providers and commissioners because, with adaptations, it could be used to aid in the development of a ‘contingency’ version of the commissioning repertoire, expressly adapted to different commissioning circumstances. This adapted version includes engagement with users (patients, public and populations), explicit concern about the quality of provision, and management–external factors such as political ‘interference’ and the media.

Personal and organisational learning and development

The action learning helped to resolve real health management problems in health-care commissioning, and through learning at a distance from work and its day-to-day pressures helped to support commissioning staff as they ‘ride the storm’. It involved reframing their management issues or problems, which opened up potential solutions not previously considered – an instance of ‘double-loop’ learning. 314 Box 1 summarises the local contexts and pressures in which the set members were operating.

We took care to distinguish the research from the action learning process. The action learning facilitator’s role was to facilitate learning, not to act as expert. Participants were encouraged to bring their own problems and identify actions to enact in the workplace. On behalf of the others, one participant summarised some of the issues they faced in their work and addressed in the learning set meetings (expressed in general terms to protect confidentiality; Box 2).

Action learning involves key interpersonal skills. The learning set supported the development of interpersonal and intrapersonal processes through encouraging and modelling active listening, careful questioning and commitment to action. Although it was self-selected, the group functioned well with high levels of engagement, trust, constructive challenge and questioning. After the final action learning set, a member who is a commissioning manager wrote a reflective piece commenting:

It has been interesting to start to apply theory to our commissioning pursuits more. Our issues are common/universal . . . It was really interesting to get an understanding of [large urban area] initiatives, where scale can be a real help . . . On the whole I think it has consolidated my view about the options open to us within commissioning, levers and incentives, service redesign versus contract and performance. I still think we need to know more about the different contractual models open to us . . . I have been struck by our continued recognition that the determinants of health are many but we are constrained in a system that is mainly health focused or works in silos for the most part. This is very much resonant with how we wish to develop an approach in [site] around using community assets and mixing health and social care interventions together.

Commissioning manager, action learning set

Also:

Reflecting on the [Livewell] partnership with a private organisation, the blurring of commissioning/providing was great and I know some people do question the split and other areas in the world have not gone in this direction. It is significantly easier to develop a pathway yourself if you have received the funding to do so and there is much untapped potential in primary care. This doesn’t translate into a strategy for how all primary care may change the way they are operating though, but does give a couple of examples of how things could be done differently.

Commissioning manager, action learning set