What does locality commissioning in Avon offer? Retrospective descriptive evaluation

Introduction

Locality commissioning is one of a range of methods whereby general practitioners can influence health services commissioning.1 In Avon it differs from fundholding and total purchasing by emphasising collaboration between practices covering geographically defined populations and by not delegating hospital and community health service budgets from health authorities to general practitioners.

Both the government and the opposition support locality approaches to commissioning, including delegation of budgets (C Smith, Health Service Journal conference, London, December 1996).2 The various models and their costs and benefits are only beginning to emerge. This study describes the consequences for general practitioners and patients and the cost to the health authority of locality commissioning without delegated budgets in one of the largest health districts in the United Kingdom.

Avon Health Authority serves 982 000 people, registered with 450 general practitioners in 147 practices (42 were fundholding in 1996-7). In 1993 general practitioners from north west Bristol formed a group to discuss locality commissioning with the health authority, and a scheme was implemented in partnership with the local medical committee and Bristol and District Community Health Council. From April 1994 all practices in the district (fundholding and non-fundholding) were encouraged to participate in 12 locality groups. Geographically close groups of practices were proposed by the authority and agreed through discussion with general practitioners. A 13th locality was added in 1996 after the local authority's boundaries were changed.

General practitioners were influential in developing the scheme. Their participation was voluntary and encouraged by the local medical committee. Representatives of more active locality groups visited and encouraged the formation of new groups. The local medical committee managed the reimbursement of lead general practitioners and practice representatives from a health authority budget. In each locality a lead general practitioner was paid to organise at least two locality meetings annually, and he or she was expected to act on issues arising. Locality groups largely set their own agendas. Agreements ranged from informal arrangements to explicitly documented agreements for new services. Locality commissioning representatives comprised five of the 13 general practitioners in the authority's general practitioner strategic advisory group. The health authority was represented in each locality group by a link person, who was a senior staff member from primary care, contracting, or public health directorates. The authority intended to develop public involvement in locality commissioning. It funded a community health council project that aimed to bring forward public opinion on locality issues.

We studied how well the objectives of locality commissioning had been met after two years and eight months, from April 1994 to November 1996. We examined evidence of service changes, costs to the health authority, time used, acceptability of locality commissioning to general practitioners, and general practitioners' perception of their influence.

Subjects and methods

We evaluated progress towards three objectives for locality commissioning that had been outlined in authority documents.

• Exploitation of local opportunities to improve health and health services

• Better understanding of local needs and priorities

• Local contribution to policy development.

• In this paper we describe the assessment of general practitioners' influence and impact (in terms of changes to services) on these commissioning objectives.

• The study was designed in consultation with the link staff and lead general practitioners. Data were collected between August and November 1996. All lead general practitioners and authority link staff were interviewed and completed a questionnaire. Interviews with the link staff were conducted by MOB, who was not otherwise active in locality commissioning. Interviews with lead general practitioners were conducted by MOB and two link staff, who did not interview lead general practitioners from their own localities. Interviews were based on an interview schedule, but interviewees were encouraged to elaborate on issues that were important to them. Link staff completed a questionnaire on each locality project and activity until the end of August 1996. Responses were validated by lead general practitioners (but not by providers). A postal questionnaire survey was conducted of a random sample of all general practitioners who were not lead general practitioners. One general practitioner was selected from each of the 147 practices by means of a random numbers table. Each respondent's answer was weighted according to the size of his or her practice to adjust for the overrepresentation of general practitioners from small practices. We examined the minutes and agendas of locality meetings.

Results

The response rates for the surveys of lead general practitioners and link staff were 100% (13/13), and for the survey of non-lead general practitioners the response rate was 74% (108/147).

Localities varied in size from five to 22 practices. Participation in locality groups was generally high but varied. Non-lead general practitioners reported that 82% (88/108) of practices had been represented at locality meetings (range 44%-100% for different localities). Lead general practitioners estimated that 82% (120/147) of practices (range 60-100%) aimed to participate in their groups. Of non-lead respondents, 25% (27/108) regularly attended locality meetings and 8% (9/108) represented their practice on a locality group but did not attend regularly. Locality groups had met an average of six times (range 2-11) during the previous year. Eleven of the 13 lead general practitioners reported that their group had planned at least one activity or project, and 10 reported having already taken action, with tangible results. Nine lead general practitioners reported acting on primary care issues, nine on secondary care issues, and six on issues to do with other agencies.

Link staff reported implementation of 20 changes to services and the planning of nine (table 1). The commonest initiatives concerned primary mental health care, nurse specialists for primary care of chronic diseases, referral and clinical practice guidelines, and access to hospital outpatient services. Locality groups initiated some changes, and they helped implement others that might have happened anyway, such as organisation of out of hours schemes. Apart from the service changes shown in table 1), we identified 29 further activities, including educational events such as a locality training day on the health needs of Somali patients, development of an information base for the locality, and investigations of issues without demonstrable changes in services to patients. External funding was not a prerequisite for changes, and several groups worked on ways of improving services within existing resources. Comments from lead general practitioners showed that the main benefits for patients were improved access to services. The main benefits to practices were improved cohesion and communication between practices, and their perception of “having a voice.”

Lead and non-lead general practitioners' attitudes to potential benefits of the scheme were assessed by the degree of agreement with statements on a five point scale (from “strongly agree” to “strongly disagree”). All lead general practitioners and two thirds of non-lead general practitioners agreed—that is, answered “agree” or “strongly agree”—with statements that communication and relationships between practices had improved (table 2). Fewer agreed that general practitioners felt more powerful or had a better understanding of how to change things. Half of the lead general practitioners agreed that current support arrangements and payment were about right for the work they did in localities. Most agreed that they enjoyed locality commissioning and that benefits outweighed the costs. The main strengths of locality commissioning identified by lead general practitioners were a flexible local approach, changes made at negligible costs, and developing larger commissioning groups to avoid “too much devolution” in fundholding.

The main barriers to greater influence identified by lead general practitioners were their lack of time and information (nine mentions), their reluctance to take on extra work (one mention), the indifference or lack of cohesion among practices (six mentions), the ambivalence about locality commissioning within the authority (three mentions), and hospital consultants' lack of understanding of primary care (one mention). To increase their groups' influence most wanted a more structured scheme with better funding (six mentions) and training (one mention). Although they were not asked their views on budgets, a few general practitioners said that they favoured budget holding as a means of increasing their influence, while one lead general practitioner wished to “steer clear of indicative budgets.” Eight lead general practitioners did not consider the scheme sustainable in its current form. Lack of time and funds were the reasons mentioned most frequently. There were comments about the scheme relying on goodwill and about the problems of lack of empowerment and visible success. They reported frustration at trying to motivate uninterested general practitioners, the slow pace of change, lack of leadership and confidence to change things, and work overload. They felt a lack of support structures and resources for their work.

Locality commissioning costs mainly comprised the time of staff in primary care, of providers in secondary care, and of the health authority. Lead general practitioners estimated that in the month before the survey they had together used 62 general practitioner half day sessions (range 0.5-8.0) on locality commissioning, which is the equivalent of just over 1.5 whole time general practitioners. Over the previous month other primary care staff used 14 sessions (0.4 whole time equivalents) and authority link staff and their secretaries used 37 sessions (0.9 whole time equivalents) on locality commissioning. The local medical committee had spent £57 687 ($92 299) on locality commissioning, which amounts to 6 p per capita. Of this, 30% was paid to lead general practitioners, 51% was paid to other general practitioners for attending locality group meetings, and 19% was spent on other items such as running meetings. Other use of time by staff of the health authority and provider units was not estimated because of the range of staff participating and because locality work could not be clearly distinguished. By November 1996 the authority had allocated £429 000 to 14 of the new service projects organised through localities. These projects were mainly primary care developments.

Discussion

This case study illustrates the potential and limitations of locality commissioning according to this model. General practitioners changed some services. Although there were no restrictions on what services could be worked on, primary care and community services were affected more than were hospital services. Most new projects arose when additional funding was available. High cost hospital services rarely encountered by general practitioners did not feature. No changes were reported for local authority services, but several localities had worked with voluntary organisations. This selective approach could reflect the early stage of development, limited financial support, or the preference of general practitioners to work only on certain issues. With greater financial and political support, general practitioners and health authorities might adopt a broader agenda, but this cannot be assumed.

General practitioner purchasers also take a selective approach. Standard fundholding is explicitly restricted. Total purchasing potentially covers all hospital and community services, but an initial survey of total purchasing sites showed that they all intended to focus on selected issues.3 Topics addressed here overlapped with those addressed by standard fundholding4 but for larger populations served by several practices.

Locality commissioning varies widely across the United Kingdom.5 6 7 The Avon scheme, in which locality groups worked directly with trust consultants and managers or reorganised services themselves, is similar to that of Newcastle and North Tyneside, in which localities were given budgets for primary care development and locality projects.7 It differs from locality commissioning in Northumberland, which emphasised general practitioners' influence on health authority contracting.8 In County Durham primary and community care needs assessment and the participation of local people were emphasised.7 Locality commissioning differs from non-fundholding schemes in Nottingham9 and Hackney,10 in each of which a single general practitioner forum worked to influence health services.

Factors influencing locality commissioning activity

Some localities were much more active than others, raising the possibility that patients in inactive localities could lose out. Activity was not consistently related to the deprivation of the population. Some of the most active locality groups were in comparatively deprived areas. The mix of fundholding and non-fundholding practices did not influence activity in a locality. Some of the most active localities had mainly fundholding practices while others had few fundholders, suggesting that locality commissioning and fundholding are complementary rather than competitive models.

A key to the scheme's successes has been the enthusiasm of participants. The doubts expressed by lead general practitioners about the sustainability of the scheme without clearer policy, more paid time, and more tangible gains suggest that this enthusiasm could decline. Limited management and funding may cause busy general practitioners to abandon the scheme while others might do comparatively little. Commissioners need the safeguards and constraints of clearer guidelines on expectations and accountability, particularly when schemes include greater financial control by general practitioners.

In Avon changes were achieved at limited additional management cost to the health authority. Development funds supported the establishment of many of the projects, but some localities pursued service improvements within existing resources. Although the time spent by general practitioners on locality commissioning seemed to exceed that funded, there was a wide range of general practitioner activity between the least and most active localities, and the previous month assessed in the survey was probably not representative of the less active early years of the scheme. Unlike standard fundholding and total purchasing, none of the management costs of locality commissioning were met with national funds. Direct comparisons with costs of standard fundholding and total purchasing were not possible as the schemes have different aims, but they seem to be greater. The direct management costs of the preparatory year of total purchasing pilot schemes ranged from £0.26 to £8.05 per capita.3 For standard fundholding, practices received £232m for staff, equipment, and computers over the first four years,4 by which time 34% of the population was covered,11 amounting to £12.90 per capita.

Methodological issues

This evaluation is different from other published evaluations of locality commissioning in terms of the duration and scope of the scheme, the range of views considered, and the inclusion of costs and consequences, but it has methodological limitations.

Firstly, we did not compare practices and populations covered and not covered by locality commissioning because the authority and the local medical committee wanted the whole district to be covered. A direct comparison with fundholding was inappropriate because most fundholders also participated in locality commissioning and the two approaches have different aims.

Secondly, information was largely based on the subjective reports and opinions of key participants. This was tempered by use of both quantitative and qualitative data and by comparing the views of lead general practitioners, non-lead general practitioners, and authority link staff, who concurred on most issues.

Thirdly, we have not measured the effectiveness of each of the changes in services that were implemented. However, three of the projects are currently being formally evaluated, and others are being monitored by audit and analysis of routine data.

Finally, a limited range of costs were measured, reflecting the difficulty in measuring indirect costs such as support from providers and the health authority.

Published reports do not allow direct comparison of costs, benefits, and limitations of these different approaches, but policy makers need to understand which combination of approaches will bring the best results and at what cost. Devolved commissioning brings additional costs, and evidence that general practitioners will take responsibility for all commissioning work is still awaited.3 Further developments in locality commissioning should be evidence based and fully costed.