Evaluation of total purchasing pilots in England and Scotland and implications for primary care groups in England: personal interviews and analysis of routine data

Introduction

The NHS Executive's initiative on total purchasing provided an opportunity for volunteer fundholding general practices throughout Britain to receive a delegated budget from their local health authority to purchase potentially all the hospital and community health services for their patients.1 We examined the factors contributing to the different levels of reported achievement among the first wave of these total purchasing pilot schemes—that is, in England and Scotland—in their first “live” year, 1996-7. We then explored the implications of the findings for the development of similar pilots proposed by the government for England—primary care groups.2

The pilots varied greatly in size, organisation, and ambition.3 As there was no detailed blueprint for total purchasing, the pilots interpreted the concept in different ways and developed their scheme at different rates. As a result, several distinct types of pilot emerged (table 1):

• Commissioners—the predominant type of pilot, characterised by holding a delegated budget and directly purchasing care;

• Copurchasers—did not hold a budget but worked in partnership with their health authority to influence its commissioning;

• Primary care developers—focused purely on primary care development;

• Developmental pilots—used the first live year as a further preparatory period;

• Undeveloped pilots—did not aim to achieve any change in services.

Methods

The diversity in approaches, combined with the fact that many objectives of the pilots were not directly quantifiable (for example, to improve interagency relations), meant that evaluation of achievement was necessarily limited to assessment of self reported progress.4 Two assessments were made: firstly, the achievement of objectives in the pilot's own terms—that is, regardless of scope—and secondly, achievements in service areas related to total purchasing. The latter assessment concerned achievements exclusively in services that total purchasers had the power to purchase for the first time: maternity services; services for seriously mentally ill patients; care of frail elderly patients in the community; accident and emergency services; emergency medical inpatient services; and the development of alternatives to acute hospital inpatient services.

There were 52 first wave pilots (46 in England, 6 in Scotland). In autumn 1995 (midway through the pilots' preparatory year) and in spring 1997 (at the end of the first live year) we conducted face to face, semistructured interviews with a lead general practitioner, project manager, and local health authority manager responsible for liaison in each project. In 1997 we asked each respondent whether the pilot had achieved its four main objectives derived from purchasing intentions documents and from the interviews in 1995. Recognising the subjective nature of the responses, and the need to reduce potential bias, we considered an objective to have been achieved only if there was agreement between all the three respondents. Moreover, where such achievements affected trusts, further corroboration was sought from provider representatives and, where relevant, hospital activity data.

Three researchers independently assigned each pilot to one of five hierarchical groups on the basis of its number of self reported achievements. For achievements in the pilot's own terms, a “group 1” pilot was one that had reported not having achieved any of its objectives in 1996-7, whereas a “group 5” pilot had reported achieving all four of its objectives. For achievements in services related to total purchasing, a group 1 pilot was one that had not influenced services in any area related to total purchasing, whereas a group 5 pilot had influenced at least three or four of these services. The grouping of pilots also took account of achievements of objectives that had emerged after the 1995 interviews. If the researchers' groupings differed, a consensus decision was reached through discussion.

Results

The figure shows the distribution of achievement reported by the total purchasing pilots in 1996-7. It shows a wide variation in the abilities of the pilots to meet their own objectives, suggesting that most had overestimated what they could achieve in the first year. For areas related to total purchasing, there was a shift towards the lower groupings, indicating that such objectives were more difficult to achieve and that a proportion of achievements in the pilot's own terms were in service areas already included in fundholding.

Distribution of total purchasing pilots by level of achievement (groups 1 to 5) in relation to objectives in pilots' own terms and to services related to total purchasing, 1996-7

• Download figure
• Open in new tab
• Download powerpoint

Table 2 shows that about half the purchasing objectives were reported to have been met (with rates varying from 39% in mental health to 87% of objectives related to extending the primary healthcare team). That the lowest rate of achievement was in mental health is scarcely surprising given the complexity of purchasing and service development in this field. In the acute sector, pilots found it easier to meet objectives relating to early discharge than to reduce emergency admissions. Moreover, objectives were more easily achieved in primary than in secondary care. The pilots wishing to influence mental health care, for example, found making changes in inpatient services that were controlled by mental health trusts more difficult than extending existing primary care services through, for example, the development of community psychiatric nursing.

Hospital episode statistics were analysed to corroborate self reported achievements for the pilots that were trying to reduce emergency admissions and lengths of stay (table 3). This analysis showed a high level of consistency between the achievements reported by the pilots and the activity changes seen in the hospital episode statistics. Table 3 shows that the hospital episode statistics corroborated the reported success in reducing length of stay in 13 of the 16 commissioner and copurchaser pilots. The other three pilots had implemented their expressed mechanism for change, but the hospital episode statistics did not show reduced average length of stay. Similar findings from the data analysis corroborate the reported achievements in influencing emergency admissions (unpublished data).

Levels of reported achievement seemed to be associated with type of pilot (table 1). Commissioner pilots tended to be the highest “achievers” in their own terms and were the only type that reported achieving all of their main objectives in services related to total purchasing. In comparison, copurchasers and primary care developers were most often medium achievers in their own terms, and most had had little or no impact on services related to total purchasing. Most developmental and all undeveloped pilots were low achievers.

Two other characteristics were associated with level of achievement1:

• Smaller pilots (those with fewer practitioners and smaller populations) were more likely than larger ones to report achieving their objectives; none of the large multipractice pilots with six practices or more was a high achiever (groups 4 or 5)

• Larger pilots had to establish complex organisations before they could make progress, whereas smaller, particularly single practice pilots, achieved their objectives with little organisational development.

The pilots with the highest direct management costs per head were the most likely schemes to meet their objectives. As the level of management spending was a matter of negotiation between each pilot and its parent health authority, the level allocated by the health authority might have been an indication of the confidence it placed in the ability of the pilot to bring about beneficial change.

Discussion

The New NHSwhite paper sets out the government's intention to establish primary care groups in England by April 1999.2 All primary care groups will have a budget covering general practitioner prescribing and practice infrastructure. As the groups develop, they will take responsibility for a greater share of hospital and community health services resources and, in the latter stages, the General Medical Services budget. Primary care groups, designed to comprise “natural communities” of about 100 000 patients, should “grow out” of the range of existing general practitioner-led commissioning schemes in the NHS. Most total purchasing pilots in England are likely, therefore, to become integrated into these larger organisations.

It is encouraging that primary care groups will be able to hold delegated budgets as we found a clear relation between first year achievements in areas related to total purchasing, and commissioner pilots. In particular, the early success of commissioner pilots in reducing length of stay shows that primary care based budget holding with independent contracting has the potential to assist in the management of demand for expensive hospital facilities. In contrast, pilots trying to influence service changes without holding a budget made far less progress in the first year.