Change in general practice and its effects on service provision in areas with different socioeconomic characteristicsBMJ 1995; 311 doi: http://dx.doi.org/10.1136/bmj.311.7004.546 (Published 26 August 1995) Cite this as: BMJ 1995;311:546
- aCentre for Health Economics, University of York, York YO1 5DD
- bHealth Policy Unit, St Mary's Hospital Medical School, University of London, London
- Correspondence to: Dr Leese.
- Accepted 7 July 1995
Objective: To investigate the changes in the structure and service provision of general practice in areas with different socioeconomic characteristics.
Design: Interview survey; postal questionnaire.
Setting: 260 group and 80 singlehanded general practices in six family health services authorities in England.
Main outcome measures: Changes in computer-isation, premises, staffing, incomes, and service provision since the introduction of the 1990 contract, including comparison with data from a study in 1987.
Results: In 1993, 94% (245) of group practices were computerised compared with 38% in 1987, and 35% (90) of practices had used the cost rent scheme since 1987. Practice managers were employed in 88% (228) of group practices, and practice nurses in 96% (249) (61% and 60% respectively in 1987). Diabetes and asthma programmes were generally more common in the more affluent areas than elsewhere. A minority of practices (27% (9/33)) in the London inner city area achieved the higher target level for cervical smear testing, compared with 88% (230) overall. A similar trend was apparent for childhood immunisation. Perceived workload increased sharply between 1987 and 1993. Differences in the mean net incomes of general practitioners between areas were much lower than in 1987. Singlehanded practices generally had more problems than group practices in improving service provision.
Conclusions: Practices in all areas have shown a strong response to the new incentives. The evidence suggests, however, that generally the urban and inner city practices still lag behind practices in rural and suburban areas in terms of practice structure and service provision.
Since 1987 increases have occurred in com- puterisation, use of the cost rent scheme, employment of practice managers and nurses, provision of disease management programmes, and health promotion activities
Urban and inner city practices still lag behind practices in rural and suburban areas with regard to service provision such as diabetes and asthma programmes
Differences in mean net incomes of general practitioners between areas are much lower than in 1987, but workload has increased
The objective of bringing all parts of the NHS “up to the very high standards of the best” has not yet been fully achieved in general practice
General practices in the more deprived parts of Britain--notably, the inner cities and urban areas--have had fewer facilities in terms of staff and premises to organise their services efficiently.1 2 The recent NHS reforms3 4 have increased pressure on such practices. Key variables in determining response in the earlier study1 2 were the local environment, size of the practice, and professional interest. There was progress but also inconsistency in general practice. Results in terms of development showed a substantial difference between the suburban and the industrial areas.
Now general practice has entered a new phase with a more powerful set of national incentives. The emphasis has shifted from local “bottom up” incentives to a “top down” contract. We examined the effect of these new incentives, which cut across the previous, slow process of development. In particular, the 1990 contract changes were national and applied to all general practitioners. The intention was that general practitioners who provided high quality services should be better rewarded financially, but few concessions were offered for those practising in areas with problems in service delivery.3 Practices in the affluent rural and suburban areas are more likely to have already the appropriate facilities and need little additional effort to supply the required services and reap the financial rewards.
Other changes, such as an increased interest in health promotion, might have taken place without the contract. The contract coincided with changed roles for family health services authorities, so impacts probably resulted from a combination of more active management and the new contract, rather than from the new contract alone.
Few studies have been published of the effects of the 1990 contract on the structure and organisation of general practice. Bain compared structural changes in a few practices,5 and the expansion of minor surgery by general practitioners has been studied.6 The use of deputising services has reduced since the new contract was introduced.7 Among general practitioners job satisfaction has decreased, mental health problems have increased,8 stress has been higher,9 and various coping strategies have been developed.10 Workload has increased.11
A pilot study in a single family health services authority in 1992 found that practices had invested heavily in equipment and services since 1986 but that differences remained depending on the location of the practice.12 Practices in the most affluent area were still more likely to invest in their premises and staff and to provide more services than those in the most deprived area, but inconsistencies in standards had been reduced.12 13
We compared in this study, however, aspects of the 1990 contract across six areas of England with differing socioeconomic characteristics and measured changes in service delivery before and after the contract was introduced. The important questions were whether the 1990 contract had improved services to patients nationally, stimulated service provision and associated issues such as staffing, and reduced the differentials between practices in different parts of the country.
Six family health services authorities or parts of them took part in the study. They were designated north west suburban; London inner city; Thames valley; east rural; north east industrial; midlands urban. They were the same areas in which our earlier study in 1987 had been undertaken.1 2 14
We telephoned one partner from all 323 group practices in the six areas to arrange an interview, with a locally based interviewer. Interviewers had attended an induction day at the University of York on using the semi-structured questionnaire and in conducting the interviews. We sent a postal questionnaire to all 142 general practitioners working singlehandedly.
Table I shows the response rates for the group practices. General practitioners in 260 of the 323 group practices took part in the study (response rate 80%; range among the six areas 71% (east rural) to 94% (north west suburban)). Response rates were better than the 74% in the previous survey,1 in which 235 group practices took part.
The practices could be divided into two distinct groups: the north west suburban, Thames valley, and east rural areas represented the affluent rural and suburban areas and were characterised by larger practices; the London inner city, north east industrial, and midlands urban areas were urban and inner city in character and had 62 of the 80 (78%) responding singlehanded general practitioners.
The six areas had been chosen to be representative of the different socioeconomic areas in England. Table II shows some of their characteristics, based on the 1991 census. The results are based on data that allow comparison over time.
In the previous study, practices were categorised as “innovators,” “intermediates,” and “traditionalists,” depending on their response to incentives.1 The traditionalists were those who had not improved their premises, not employed a practice nurse, nor opted to take part in the vocational training scheme. Since 1987, the traditionalists have virtually disappeared (table III). An important shift in practice organisation has taken place in all areas. Computerisation, use of the cost rent scheme (government subsidy for improving general practice premises), and employment of practice managers and of practice nurses have all increased rapidly. The number of training practices, however, has changed little since 1987 (a rise from 74 to 80).
Table IV shows that, overall, 94% (245) of group practices had a computer (85% (London inner city) to 100% (north east industrial and east rural)). In 1987, 36% of group practices had a computer.2 However, although computer ownership was high in all six areas, the computers were used differently (table IV). Overall, 19% (49) of group practices were still setting up their computer system (30% (London inner city) to 3% (east rural)). In all, 48% of group practices purchased computers during 1988-90 (table IV), and many did so around the time that the 1990 contract was introduced.
In total, 91% (236) of practices had computerised age-sex registers (76% in the London inner city area, reflecting its lower level of computerisation) (table IV). In 1987, only 46% of practices in the midlands urban area had an age-sex register, compared with 67% in the London inner city area and 75% overall. In this study, computer use for health promotion data also varied by area, the lowest level again being in the London inner city area (67%), compared with 84% overall (P<0.01) (table IV).
Cost rent scheme
Table V shows the number of group practices that have made improvements under the government's cost rent scheme since 1987 (table V); pressure on space had led to a new drive to investment. Overall, 35% of practices had used the cost rent scheme since 1987 (24% in the midlands urban area and 31% in the Thames valley area).
Before 1990, 65% of group practices overall had a practice manager (48% (London inner city) to 83% (east rural)), compared with 88% in 1993 (76% (midlands urban) to 98% (east rural)) (table VI). Table VI also shows the increase in the number of practices with a practice manager and whether practices had practice or primary health care team meetings (most do, at least occasionally).
Table VI shows that most group practices currently employed a practice nurse (96%) with little difference between areas. Before 1990, 78% of practices employed a practice nurse (73% (London inner city) to 88% (east rural)). In the 1987 study, 60% of practices employed a nurse (42% (midlands urban) to 75% (Thames valley)).
SERVICES FOR PATIENTS
As part of the 1990 contract, general practitioners were to be paid for providing specified disease management programmes for patients with diabetes or asthma. Table VII shows that the uptake of such programmes has been high in all areas. Diabetes is more common, however, among elderly people and people from ethnic minority groups,15 both of whom are more likely to live in the more deprived areas of inner cities. Although the areas studied differed little in the proportion of practices providing these services for asthma and diabetes, the urban areas had generally lower levels of provision than the more affluent rural and suburban areas (P<0.05).
Most practices achieved high health promotion bands (table VII). Overall, 93% of group practices were in band 3 (the highest band, for which practices receive the highest fee), with the highest proportions being in the north west suburban and east rural areas (100% and 95% respectively). However, although, overall, 3% of practices were in band 2, the corresponding proportions for practices in the London inner city and the midlands urban areas were 9% and 7% respectively. In the London inner city area 6% of practices were not allocated to a band at all (2% across all practices), with no unallocated practices in the north west suburban and east rural areas.
Table VIII shows the target levels achieved for uptake of cervical smear testing, immunisation of children under 2 years, and immunisation of children under 5 years in each area. For cervical smear testing, 88% of practices achieved the higher level (27% (London inner city) to 100% (north west suburban). A similar trend was apparent for immunisation of children under 2 years--94% overall (73% (London inner city) to 100% (east rural)). Of the 25 practices that only achieved the lower level target for immunisation of children under 5 years, 23 were in the more industrial areas.
The 1990 contract introduced fees for minor surgery. Here differences between areas were rather greater (table IX). In all, 80% of general practitioners were on the minor surgery list, and 84% of practices that were asked offered minor surgery (76% (London inner city) to 98% (east rural); P<0.01). Lower proportions of general practitioners were found on minor surgery lists in the London inner city (51%), midlands urban (71%), and north east industrial (83%) areas than in the east rural (88%), Thames valley (83%), and north west suburban (85%) areas (P<0.01). Provision had increased since the 1990 contract in all areas but most in the London inner city and north east industrial areas.
The contract provided an additional capitation payment for child health surveillance services to children under 5 years. Table IX shows the numbers of general practitioners on the child surveillance list by area; although the proportion overall was 68%, the proportions varied widely among the areas (42% (north west suburban) to 88% (north east industrial)). Practices may have decided that not all their general practitioners should be on the list and that the service should be provided by specified general practitioners only. These data do not therefore show that only 68% of all practices provide this service.
General practitioners' perceptions of their workload changed dramatically between 1987 and 1993. Those who were able to cope fell from 43% to 16% in 1993, while those under great pressure increased from 12% to 40%. Those under some pressure had remained fairly constant.
The new contract and the new pay system have changed workload and increased stress. They have also had appreciable effects on differences in pay between areas. Differences in average net incomes were much smaller in 1993 than in 1987 (table X). In particular, net incomes had risen in socially deprived areas. As a result, areas were much closer to the review body's average in 1993 than had been the case in 1987. As the data in table X show, incomes have levelled out in relation to the review body's targets.
In all, 56% (80/142) of singlehanded general practitioners responded to the postal questionnaire (49% (24/49) in the midlands urban area to 75% (3/4) in the east rural area). Two practices, however, were excluded from the study because they had too few patients (100 and 16), leaving 78 practices for analysis.
Computerisation varied more among singlehanded general practitioners than among group practices. All practices in the east rural (3) and north east industrial (18) areas were computerised compared with 58% (11/19) in the London inner city area and 63% (15/24) in the midlands urban area. Of all singlehanded general practitioners, 76% (59) had a computer.
In all, 60% (47) of singlehanded practitioners employed a practice manager in 1993 compared with 22% (17) who had done so before 1990. All practitioners in the north west suburban (4) and east rural (3) areas employed a practice manager; the corresponding proportions for the other areas were 63% (12/19) in the London inner city area, 60% (6/10) in the Thames valley area, 50% (9/18) in the north east industrial area, and 54% (13/24) in the midlands urban area. Single-handed practices were therefore much less likely than group practices to have the support of a practice manager and much more likely to be in the urban and inner city areas. They were also much less likely to hold primary health care team meetings (49% (38)), although 82% (64) held practice meetings.
Fewer singlehanded general practitioners than practitioners in group practice employed a practice nurse. Overall, 85% (66) employed a nurse (only 68% (13/19) in the London inner city area and 70% (7/10) in the Thames valley area). By contrast, all 21 practitioners in the east rural and north east industrial areas employed a practice nurse.
Among singlehanded general practitioners, 79% (62) had approval for asthma management programmes and 83% (65) for diabetes management programmes, with the lowest rates in the London inner city area. In all, 81% (63) of singlehanded general practitioners were in band 3 for health promotion, with 71% (17/24) in the midlands urban area and 74% (14) in the London inner city area. Nine per cent (7) of practices were not allocated to a band (16% (3/19) of practices in the London inner city area and 17% (4) of practices in the midlands urban area). For cervical smear testing 77% (60) achieved the higher target level. The proportion achieving the higher level in the London inner city area (37% (7/19)) was much lower than in the other areas (80% (8/10) in the Thames valley area and 100% in the north west suburban (4), east rural (3), and north east industrial (18) areas).
For immunisation of children under 2 years, 85% (66) of practices achieved the higher target level and 12% (9) the lower. But the higher level was reached by 68% (13/19) of practitioners in the London inner city area compared with 80-100% in the other areas. Similarly, for the children under 5 years, 81% (63) of practices achieved the higher target but only 68% (13/19) in the London inner city area and 72% (13/18) in the north east industrial area did so, compared with 88-100% in the other four areas. In all, 69% (54) of practitioners were on the minor surgery list (but only 37% (7/19) in the London inner city area).
The results show that singlehanded general practitioners were having more problems than group practices in responding to changes and increasing activity; but their responses were highly variable between areas. The performance of practitioners in the London inner city area influenced targets and activity across the area; but practices in the midlands urban area showed a stronger response despite problems of relating to patients from different ethnic backgrounds.
The combined incentives operating since 1987 have brought about intensive changes. The previous pattern of change based on differences of area response to incentives has been overtaken. Instead, practices in all areas and of all sizes have shown a definite and strong response to the new incentives.
The evidence suggests, however, that in general the urban and inner city practices still lag behind practices in rural and suburban areas. This is particularly true for the London inner city area, where practices have many disadvantages.16 Practices in this and similar areas have greater difficulty in providing the new services demanded by the 1990 contract. These are also the areas of high morbidity and deprivation. The population of the London inner city area is characterised by severe overcrowding, the highest representation of ethnic groups of all the study areas, and high levels of deprivation, shown by the underprivileged area score.17 This area also had a high proportion of singlehanded general practitioners and large list sizes, although these were greater in the midlands urban area. Despite the two areas' similarity on socioeconomic factors, the midlands urban area scored more highly than the London inner city area on targets achieved for immunisation and cervical smear testing. High levels of deprivation are not therefore in themselves a barrier to the provision of the health promotion services required by the 1990 contract. One factor, however, that might have led to apparent under-performance in the London inner city area is the possible presence of “ghost” patients--namely, patients registered with a practice but who have moved. They would appear on practice lists but would not be receiving services.
The results suggest that in the past five years general practices have moved towards more uniformity, with much greater workload and less variation in net incomes. The new contract has generated more activity and changed process. The new combination of incentives--partly income, partly managerial--has led to increased activity across all areas. National provision has brought about an increased rate of change in all areas without eliminating local variations. The objective of bringing all parts of the NHS “up to the very high standards of the best”18 has not yet been fully achieved in general practice. Practices are working against a moving target: they may achieve standards that would have been considered adequate at one point in time, but innovators continue to move ahead.
We thank Gillian Rodrigues for help with the data and Dr Nigel Rice for helping in the statistical analysis. We also thank our local interviewers and the general practitioners who gave their time so generously.
Funding Economic and Social Research Council.
Conflict of interest None.