Article Text

Download PDFPDF

The work of commissioning: a multisite case study of healthcare commissioning in England's NHS
  1. Sara E Shaw1,
  2. Judith A Smith2,
  3. Alison Porter3,
  4. Rebecca Rosen2,
  5. Nicholas Mays4
  1. 1Queen Mary University of London, London, UK
  2. 2Nuffield Trust, London, UK
  3. 3College of Medicine, Swansea University, Swansea, UK
  4. 4London School of Hygiene and Tropical Medicine, London, UK
  1. Correspondence to Dr Sara E Shaw;{at}


Objective To examine the work of commissioning care for people with long-term conditions and the factors inhibiting or facilitating commissioners making service change.

Design Multisite mixed methods case study research, combining qualitative analysis of interviews, documents and observation of meetings.

Participants Primary care trust managers and clinicians, general practice-based commissioners, National Health Service trust and foundation trust senior managers and clinicians, voluntary sector and local government representatives.

Setting Three ‘commissioning communities’ (areas covered by a primary care trust) in England, 2010–2012.

Results Commissioning services for people with long-term conditions was a long drawn-out process involving a range of activities and partners. Only some of the activities undertaken by commissioners, such as assessment of local health needs, coordination of healthcare planning and service specification, appeared in the official ‘commissioning cycle’ promoted by the Department of Health. Commissioners undertook a significant range of additional activities focused on reviewing and redesigning services and providing support for implementation of new services. These activities often involved partnership working with providers and other stakeholders and appeared to be largely divorced from contracting and financial negotiations. At least for long-term condition services, the time and effort involved in such work appeared to be disproportionate to the anticipated or likely service gains. Commissioners adopting an incremental approach to service change in defined and manageable areas of work appeared to be more successful in terms of delivering planned changes in service delivery than those attempting to bring about wide-scale change across complex systems.

Conclusions Commissioning for long-term condition services challenges the conventional distinction between commissioners and providers with a significant amount of work focused on redesigning services in partnership with providers. Such work is labour-intensive and potentially unsustainable at a time of reduced finances. New clinical commissioning groups will need to determine how best to balance the relational and transactional elements of commissioning.

  • Health Services Administration & Management
  • Qualitative Research

This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 3.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Article summary

Article focus

  • Commissioning—or strategic planning and purchasing—is central to current reforms of the English NHS, which aim to strengthen the role of clinicians in commissioning and the use of market forces.

  • Little is currently known about what ‘effective commissioning’ is and how it can be achieved in practice.

  • This study examines the work involved in commissioning long-term condition services and considers the factors inhibiting or facilitating commissioners in making service change.

Key messages

  • Commissioning for long-term condition services challenges the conventional distinction between commissioners and providers, with a significant amount of work to review and redesign services undertaken in partnership with providers.

  • There is little evidence of commissioners using market-style elements of commissioning, such as decommissioning or tendering for new forms of service, in planning and purchasing long-term condition services.

  • The scale and intensity of work that is undertaken to commission long-term condition services appear to be disproportionate to the likely service gains. Decision-makers need to think differently about how to commission long-term condition services.

Strengths and limitations of this study

  • The strength of this study lies in the detailed examination of the day-to-day work involved in planning and purchasing long-term condition services, and the level of engagement and development work that this reveals, often in partnership with providers and other stakeholders.

  • Given the emphasis in current NHS reforms on extending market-style transactions, the study findings raise timely questions about the operation of a healthcare market in the NHS and, in particular, about the value of a clear split between commissioners and providers of healthcare.

  • The study did not set out to analyse the costs associated with commissioning work. However, the level of work involved in commissioning compared to the likely service gains reveals it to be an area deserving of closer examination.


Commissioning is a term used in the English National Health Service (NHS) to refer to a proactive and strategic process for the planning, purchasing and contracting of health services.1 Effective commissioning is regarded by NHS policymakers as crucial to achieving high quality care that is responsive to patients’ needs and ensures value for money.2–4 Little is known about how effective commissioning can be achieved in practice.

This paper examines the work of healthcare commissioning. It builds on recent research examining the way that commissioning is understood and undertaken locally,5–7 the spaces in which commissioning takes place8 and the people involved.9 ,10 The focus is on the commissioning of services for people with long-term conditions. Findings are reported from a multi-site case study of NHS commissioning in England, aiming to identify the ways in which commissioning is enacted and the factors inhibiting or facilitating progress in making service change.

Findings are relevant to those health systems that have introduced healthcare commissioning, and are particularly pertinent to recent reforms to the NHS in England.4 The aim of these reforms is: to strengthen the role of clinicians in commissioning; support patient-centred care; enhance the quality and diversity of providers1113 and increase the reliance of commissioners on competitive tendering and other market mechanisms,4 ,14 thereby reducing the salience of the more relational aspects of commissioning (such as collaborative service planning).15–18 Previously, the Department of Health had promoted an annual process of needs assessment, planning, contracting, monitoring and review, often referred to as the ‘commissioning cycle’ (figure 1, adapted from the Department of Health,19 following Ovretveit20). As the organisations responsible for commissioning local healthcare up to April 2013, primary care trusts were encouraged to follow this annual process. From April 2013, 211 clinical commissioning groups led by general practitioners (GPs) have taken on similar roles and responsibilities in relation to commissioning.

Figure 1

The commissioning cycle.


We conducted a case study of three ‘commissioning communities’ (the area covered by a single primary care trust, table 1) in England. Each site replied positively to an invitation sent to primary care trusts identified as performing better (for instance, in relation to ratings of service quality, resource use) than would have been expected when compared with similar organisations (see final report for details21). Each commissioning community included primary care trusts, clinical commissioners, hospitals, community and mental health service providers, local government and the independent sector (table 1). During the research, the Calderdale case study was extended to include neighbouring Kirklees, reflecting close partnership working.

Table 1

Overview of commissioning stakeholders included within case studies

The three communities served populations of 200 000 (Calderdale), 525 000 (Somerset) and 310 000 (Wirral). Spending on healthcare was similar to the English average.22

The study comprised four phases and data collection within each phase is detailed in table 2. The findings presented here draw largely on phase 3 in which observations and interviews were undertaken across the three sites between November 2010 and January 2012. This was supplemented with an analysis of national and local documents relevant to each of the conditions and commissioning communities studied.

Table 2

Phases of the study and data collected

The study focused on two long-term condition services in each of the three sites. Diabetes was selected as a condition across all the three sites. Each commissioning community then identified a second long-term condition on which they wished the research to focus: dementia in Calderdale and Wirral, and stroke in Somerset. Each of these long-term condition services was written up as a descriptive account (up to 65 pages), which was amended as new data were collected. We then undertook thematic analysis23 and examined connections between the inputs (people, organisations, data, money, ideas and time) and processes (driving change, addressing local needs, specifying services and agreeing contracts, measuring and promoting service quality and reviewing services). We combined this with indicative coding, ensuring that we identified issues not anticipated in initial research questions but with implications for healthcare commissioning. We examined emerging themes within each case and then compared commissioning practices across the three communities to identify variation, as well as those aspects of commissioning that produced changes in the way in which services were provided (including enhanced clinical effectiveness, as well as other purposes such as cost containment).


In relation to diabetes, we studied the commissioning work allied to the development of a strategic plan (column 1, table 3), development of a new model of diabetes care (column 3, table 3) and review of a diabetic podiatry service (column 5, table 3). We also studied a plan for major changes to dementia services (column 2, table 3), establishment of a new early supported discharge service for stroke patients (column 4, table 3) and establishment of a new memory assessment service (column 6, table 3).

Table 3

Overview of selected long-term condition services*

Each of the six areas studied undertook commissioning activity that entailed review and redesign one or more aspects of service delivery for long-term conditions. This work was driven by a range of local factors, including a need to address rising local prevalence by increasing the capacity and/or accessibility of services (columns 2, 3 and 6, table 3) and an aspiration to develop a new model of care (columns 1, 4 and 5, table 3). Commissioning work tended to be driven by a local or national push for service review. In one instance, the diabetic podiatry service in Wirral (column 5, table 3), the service was a long-standing local concern.

We intended to study a single annual commissioning cycle in each of the six service areas. It quickly became apparent that the commissioning process for long-term condition services did not fit neatly into a single year and involved a range of activities that were not typically thought of as a part of the commissioning cycle, including convening and coordinating service development across interest groups and supporting service implementation (columns 2–4 and 6, table 3).

Progress within each of the six service areas was varied. Two services remained in the early stages of the commissioning process due to limited commissioning staff capacity (1 and 2); one service developed further as a result of progress with a new computer system (5); and three new services were successfully launched following several years of planning (3, 4 and 6).

The commissioning work that we observed was complex and multifaceted, involving effort by a wide range of individuals and organisations and taking place over long periods of time. Any resultant changes in the provision of care tended not to be as great as the commissioners had hoped for. To examine the organisation and processes contributing to effective commissioning, we focused on five areas:

  • The process of commissioning

  • The type of activities undertaken

  • The range of people involved

  • The time and effort expended

  • The potential service gains allied to commissioning

Commissioning for long-term conditions is not a neat and sequential process

The annual commissioning cycle (figure 1) was regarded by participants as a useful model for making sense of commissioning work, but in reality, activity rarely followed this neat, annual cycle.

Once an area of commissioning work had been identified, activity typically stretched over several years, with starting points dating back as far as 2007 (table 3). Early development work was particularly time-consuming: It takes years and years to do anything and…you've got to wait for the next meeting and another month for that and another month for this [Clinical commissioner].

A minimum of 1 year was typically spent assessing needs, reviewing evidence and developing the service specification. Public health data were used to support and legitimise emerging commissioning plans, rather than drive them from the outset. Once a firm decision was made to move ahead with service redesign—as with the diabetes plan and early supported discharge service in Somerset (columns 3 and 4, table 3) and memory assessment service in Wirral (column 6, table 3)—progress seemed to speed up, and the service model, referral procedures and staffing were established within months rather than years.

Commissioners judged success largely in terms of whether the service was running smoothly and efficiently (ie, activity levels in relation to cost). There was less emphasis on whether the right delivery model was in place. This reflected a tendency across sites not to engage in discussions about discontinuing or replacing services. As one Primary Care Trust (PCT) senior executive put it, “I'm not sure the NHS has a good history of reviewing services in that way.”

Across all six areas, only one process involved decommissioning an existing service model, with the memory assessment service in Wirral (6) replacing a memory clinic run by GPs with a special interest in dementia.

Commissioning services for people with long-term conditions in the NHS is highly relational

Commissioning is increasingly envisaged in NHS policy4 as a predominantly transactional process, whereby commissioners select providers competitively and contract with them to deliver a specified service. However, we observed that the bulk of work carried out by commissioning staff involved collaborative activities. These included: working to build consensus and address priorities; gaining input from providers and other stakeholders (including patients) on specific aspects of service plans and managing change associated with implementing new services.

In the three service areas that achieved the most progress in terms of service change in the direction proposed by commissioners—the diabetes service (column 3, table 3) and early supported discharge service (column 4, table 3) in Somerset and the Wirral memory assessment service (column 6, table 3)—relational work related mainly to strategic leadership, involving the identification of clear priorities and ensuring that there was commitment on the part of local providers, clinical staff and other interest groups. In all 6 areas, implementing service change was an integral part of commissioning work, with the emphasis on facilitation: [the] key person that's able to coordinate efforts across everybody and actually just keep on, keep saying ‘Right we've got another meeting…have we done what we said we were going to do?’ [Primary care trust manager].The more transactional aspects of commissioning came into play when a deal needed to be struck in respect of finance and contracts. Discussions about funding and contract negotiation were particularly sensitive, tended to take place behind closed doors and outside mainstream commissioning work and were less accessible to the study team. In contrast to relational work, the management of contracts tended to operate in line with an annual commissioning cycle with staff working to fixed deadlines.

All three sites separated the negotiation and management of contracts from strategic development and service redesign work, with different staff involved in the two types of roles. Contracting appeared to be a largely transactional process, involving clear timescales and processes and with defined roles for commissioners and providers. Interviewees described how contracting work tended to be undertaken by a small group of specialised commissioners, many with financial expertise. However, transactional commissioning was described as depending on prior relational work, with flexibility and reciprocity being crucial in maintaining momentum for change, particularly given the increased demand for long-term condition services and potential financial shortfalls.

Providers play an important role in commissioning

The official model of commissioning promoted by the English NHS describes commissioners as those who plan and fund services to meet local healthcare needs, clearly distinct from those who provide services. However, the tasks of commissioning were not carried out exclusively by people with the title commissioner in their job description. Managers and professional staff from provider organisations and local authorities, clinicians and, to a lesser extent, patients and the third and independent sectors also played a role. Contributions varied at different stages of the commissioning process with, for instance, service user input being more prominent in the planning stages.

GPs (particularly those who were also involved with local practice-based commissioning initiatives56) and NHS hospital, mental health and community health service providers took a particularly active part in commissioning. The principle of active partnership across commissioners and providers was fundamental to discussions about healthcare needs and service design, as well as to developing approaches to service monitoring: it's very much a collaborative, inclusive process that then produces the model of service and also [considers] affordability [Senior executive, acute/mental health provider]

In three cases (columns 2, 3 and 6, table 3), providers took a lead role in commissioning, bringing specialist knowledge of clinical care and specific skills in project management, coordination and leadership. This leading role was regarded positively as “partners helping each other work with situations” rather than “adversaries trying to screw every last advantage out of each other” (Manager, provider organisation). However, a clear distinction was made between contracting—where distance between the commissioner and the provider was considered essential—and more relational aspects of commissioning where partnership working across the purchaser-provider split appeared to be the norm. As one primary care trust manager reported: It's not…a cosy relationship. It can't be, because it's…also got, you know, a business function. You are there to assure…the organisation within which you sit, wherever you're a commissioner—and ultimately the Board and you know, at a national level—how you are making best use of public money.

Individual doctors and other staff from local providers contributed positively to service planning. Clinical staff were highly valued by commissioners, enabling them to publicise potential service changes to the wider clinical community.

Commissioning long-term condition services involves intensive labour

Much of the work of commissioning across all 6 service change processes was focused on service development. This tended to be concentrated on small areas of service provision, and appeared to be labour-intensive and time-consuming.

A significant amount of work involved commissioners convening wide-ranging groups of people over whom they had little—if any—managerial authority. The focus of this work was on developing and sustaining strategic partnerships as a routine part of commissioning. This coordinating role was most visible in diabetes (column 3, table 3) and early supported discharge (4) services in Somerset and the memory assessment service in Wirral (6) where partners described how they had “always worked together”, and how commissioners had “always sought their view on service delivery”.

We observed an extraordinary amount of effort going into the relational aspects of commissioning and, in particular, to establishing, running and managing formal meetings allied to the service development work of commissioning: If you think again just in terms of the timeline, you know, all the meetings that were involved, this took people away from other things. And the work involved in writing up papers, doing the presentations, struggling with putting together a programme [PCT manager].

Meetings ranged from one-off events (eg, a workshop on transforming dementia services in Calderdale, involving over 80 stakeholders) to regular planned meetings (eg, Wirral Older People's Services Network, a regular joint strategic planning meeting involving commissioners, providers and service users; table 4). The majority were led by commissioners, requiring considerable managerial and administrative time and extensive participation of clinical and non-clinical stakeholders.

Table 4

Overview of meetings in the commissioning process

A similar picture was evident in relation to other commissioning tasks, including: needs assessment, evidence review, demand mapping, modelling, designing care pathways, writing service specifications, preparation of business plans and developing outcome measures. All were essential but time-consuming parts of the commissioning process. None were observed (or reported) as taking place in relation to specific phases of the commissioning cycle. For example, commissioners told us that they placed a high value on using data to support evidence-based decision-making. However, in practice, the task of collecting and reporting up-to-date data was onerous with data systems often being incompatible between providers or inadequate to the task there's a consistent problem about systems and repositories and data and how you share it [Manager, local government]

Inconsistent categorisation of activity (eg, not recording diabetes as a secondary diagnosis) compounded problems with accessing data.

The scale and intensity of commissioning work may not always be proportionate to the impact

The scale and intensity of the commissioning work that we observed led us to examine what was being secured through this work. Across all sites, the scale and intensity of work often appeared to be disproportionate to the anticipated or actual service gains.

The three service areas that made the most progress with remodelling services (columns 3, 4 and 6, table 3) required considerable labour to develop long-term condition services. However, each had also adopted an incremental approach to commissioning, and to change more broadly, which appeared to enable them to keep the labour more manageable and focused over time. This approach was described to us as ‘intelligent commissioning’ (6), ‘staged development’ and ‘learning in practice’ (4) and was characterised by planned evolutionary change; a large-scale vision for the specific long-term condition service (including linking with national priorities and guidance), combined with focused and actionable tasks; senior managers with capacity and support to lead change; partnership working characterised by trust, as well as mutual challenge; and focused collection and use of data to guide and support decisions. The Somerset diabetes services (3) and the Wirral memory assessment service (6) were also characterised by ongoing review and negotiation to match finance to demand, which worked well for both partners So it's a vicious circle if you like, because the more staff that we have, if we can find the funding for those posts, the more assessments they can undertake, and they may well lead in to more people needs, you know, on-going treatment and prescribing. So there are some commissioning, ethical discussions to be had about how we move that forward. [Primary care trust senior executive]

Those developing the diabetes (1) and dementia (2) services in Calderdale and the diabetic podiatry service in Wirral (5) also expended considerable labour. However, they struggled to focus their work, to find capacity to identify and pursue actionable tasks and to bring about change through commissioning. In Calderdale, for example, commissioning staff were described by one primary care trust senior executive as “stretched, absolutely stretched”, requiring them to focus on service areas other than dementia and diabetes. This was compounded by difficulties in identifying appropriate units of commissioning work (ie, ‘projects’), which needed to be big enough to justify the work involved, while remaining manageable.

Services for stroke and diabetes in Somerset (3 and 4) and the Wirral memory assessment service (6) struck this balance well, working with existing services and structures, and alongside providers, to focus on manageable areas of activity (table 3). In contrast, ambitions for large-scale ‘transformation’ of diabetes and dementia care in Calderdale were hampered by a lack of focus our Mental Health Trust actually came up with the idea of really looking at the dementia pathway and doing some significant work on it and [then] there were a number of enablers across all long term conditions that would support people with dementia as well as people with other long term conditions such as supported decision making, telehealth, predictive risk, all the sort of things in the system, generic workers, community matrons needed to be in place. [Primary care trust manager]

The focus on diabetic podiatry in Wirral (5) appeared to involve considerable effort devoted to a small area. With diabetic podiatry cutting across several areas of commissioning work (for eg, diabetes, community podiatry and emergency foot care), those involved in commissioning appeared to find it hard to look beyond the multiple and complex connections across these areas and focus on specific and manageable projects.

Commissioners documented and discussed the cost of delivering services and anticipated gains from commissioning. In the short term, they anticipated benefits in the quality of care within each of the six service areas (eg, reduction in amputations due to improved diabetes care in Wirral). In the longer term, potential savings were thought likely to accrue over a period of 5–10 years through substitution (for eg, with an increasing level of low risk foot care undertaken by nurses and healthcare assistants in general practice), reductions in hospital admissions (particularly for dementia and stroke) and/or assisted living in the community (for instance, increasing the number of people with dementia able to live at home for longer). However, while commissioners clearly aspired to benefits in quality and efficiency, there was little indication of what savings might realistically accrue from their work.


This study has revealed the multiple and labour-intensive processes associated with commissioning. While the commissioning cycle (figure 1) provided a useful guide for primary care trusts, commissioning activities did not follow a neat series of stages within an annual cycle. At least for long-term condition services, commissioning involves an evolutionary process of service review and redesign, often spread over several years, and in partnership with providers and other stakeholders. This process involves an extraordinary amount of work and it remains unclear if this is worth the likely impact. Money and resources appeared to feature infrequently in commissioning discussions, with little assessment of the cost of commissioning work or the likely cost-effectiveness of proposed service developments.

Our study focused specifically on the work involved in commissioning long-term condition services. We were not able to directly observe the more contractual—or transactional—elements of commissioning discussions which appeared to take place elsewhere. However, our analysis of interviews and documents—as well as observation of the day-to-day activities involved in commissioning—confirmed that commissioners tend to focus on the relational rather than transactional aspects of commissioning. This was evident in the time and energy given to consultation, planning and review meetings and each PCT's role in coordinating the local healthcare system. The use of contracts and funding to bring about change tended to be divorced from, or seen as less important than, these wider commissioning activities, suggesting that commissioners were not entirely comfortable with the more transactional elements of their role involving, for instance, decommissioning services or seeking new alternative providers.

Our focus on the work of commissioning, as well as on the processes that make up the commissioning cycle (figure 1), makes this study distinctive. Previous research on commissioning has tended to focus on how national policy facilitates or inhibits effective commissioning11 ,24–29; the organisation of commissioning30–34; and specific aspects of the commissioning cycle,35–38 such as contracting or procurement. This research adds to the literature, focusing on the detail of commissioning practice and revealing activities that seem to contribute to more effective commissioning (in terms of service change in the direction proposed by commissioners). Commissioners developing a new model of diabetes care for Somerset (3) and the memory assessment service in Wirral (6) mapped out a coherent programme of commissioning for each service, linking this with strategic priorities and funding, striking a balance between relational and transactional activities, and making change in a way that enabled the new service to develop at some scale.

Research on the nature of contracts for healthcare has identified the importance of ‘relational contracting’,17 ,18 where trust between the parties can help mitigate difficulties associated with the absence of complete contracts. Our findings extend this concept of ‘relational contracting’ to the wider commissioning function, drawing attention to the scale and intensity of labour expended. This resonates with recent research demonstrating the extent of the partnership working characteristic of commissioning processes for services for people with long-term conditions.39 It also supports recent research describing commissioners as ‘animateurs’, attempting to bring together and influence a disparate group of people over whom they have little direct managerial authority.5

Our research suggests that, at least for long-term condition services, decision-makers need to think differently about the way in which commissioning is carried out and about the operation of a healthcare market. The findings show that commissioning tends to be a labour-intensive process often undertaken in partnership with providers, blurring the distinction between commissioners and providers emphasised in the recent policy.4 The amount of work and extent of partnership working required remains open to debate. However, it is clear—from our findings and the wider literature—that commissioning (and contracting) can neither be undertaken by transactional means alone, nor indeed by purely relational activities. The findings therefore challenge the value of a clear split between commissioners and providers of healthcare (a key organising principle of the NHS quasi-market for over 20 years32) in all situations and all stages of the commissioning process.

In a publicly funded healthcare system—with goals of value for money and equity of access and outcomes—there is inevitably a need for some sort of commissioning or planning function to decide how much to spend on which services and with what aims. Our study has enabled a detailed examination of this process. It has revealed that commissioning services for people with long-term conditions appears to be characterised by a predominance of relational commissioning, with little evidence of commissioners using the ‘harder’ elements of commissioning practice (such as tendering for new forms of service).40 This raises a question as to how the NHS can best direct commissioning work, particularly at a time of reduced management costs.41 Our research did not include analysis of the costs associated with commissioning work but has revealed it to be an area deserving of closer examination in future. In the NHS, choices will need to be made as to how much engagement and development work commissioners will be able to do in the future and, like any managerial activity, what are the most efficient ways of doing commissioning. Clinical commissioners will need to determine how best to balance the relational and transactional aspects of commissioning: encouraging providers to take a lead role in service development and redesign and so help to fill the gap left by the limited capacity and resources; bringing money (and value for money) to the fore in commissioning discussions; using contracts in a more focused way and exploring opportunities for reviewing, discontinuing and re-commissioning services.


The authors thank all of those who participated in the study, the research advisory group and colleagues at the Nuffield Trust and London School of Hygiene and Tropical Medicine, whose input and expertise have been invaluable throughout.



  • Contributors SS, JS, RR and NM contributed to the research proposal, applied for the NIHR grant and conceptualised the study. SS applied for the ethical approval. SS, AP and JS completed the data collection. JS and RR undertook the support work with the commissioners. All authors contributed to the analysis of the data and take responsibility for the integrity of the data and the accuracy of the data analysis. SS conceived of the paper and JS, AP, RR and NM contributed to the successive versions. JS was the principal investigator and is the guarantor for the paper.

  • Funding The project was funded by the National Institute for Health Research Health Services and Delivery Research Programme (project number 08/1806/264). The views and opinions expressed in this paper are those of the authors and do not necessarily reflect those of the National Institute for Health Research Health Services and Delivery Research Programme or the Department of Health. The funders were not involved in the selection or analysis of data, or in contributing to the content of the final manuscript.

  • Competing interests The authors have support from the National Institute for Health Research for the submitted work.

  • Ethics approval The study received ethical approval from the NHS Outer South East London Research Ethics Committee (reference number: 09/H0805/40). All those participating in the study gave their informed consent before taking part.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement No additional data are available.