What will a primary care led NHS mean for GP workload? the problem of the lack of an evidence base

Background

A primary care led NHS places emphasis on shifting the balance of care from the acute hospital sector to primary care.1 2 3 This is only one of the many changes taking place in the NHS, and its effect on workload in primary care is potentially far reaching.

The trend for more services to be provided in primary care has been in progress for some time, principally as a result of technological changes and rising consumer demand. Furthermore, the 1990 general practitioner contract created financial incentives for general practitioners to replace some hospital based services with practice based provision, for example, by providing services such as minor surgery and chronic disease management for diabetes and asthma.4 Fundholding created incentives for general practitioners to provide secondary care in their practices, for example, by using savings to invest in their premises and practice based facilities.5 Further developments included in guidelines from the NHS Executive in 1993 allowed fundholders to use their budget to pay either themselves or other health professionals to provide a specified list of secondary care services.6

Changed discharge practices in the acute trusts means that general practitioners and other practice staff or community nurses are likely to be caring for elderly people discharged from long stay and acute inpatient hospital departments

BRENDA PRINCE/FORMAT

• Download figure
• Open in new tab
• Download powerpoint

Other changes that had implications for general practitioner workload, such as long term policies to move care away from institutional and hospital settings towards care in the community, were re-emphasised in the 1990 NHS and Community Care Act.7 8 In addition, the internal market encouraged trusts to change discharge practices by reducing length of stay in order to maintain or increase throughput, in response to financial pressures created by the purchaser-provider split.

The NHS Executive's publication which set out “a national framework for the provision of secondary care within general practice”9 also had implications for general practitioners' workload. This trend has accelerated with the changes put forward in Primary Care: The Future and in Choice and Opportunity.10 11 Although the reforms offer general practitioners a central role in the NHS, there are reports of increasing dissatisfaction,12 13 14 coupled with substantial resistance to change, with general practitioners pointing to anecdotal evidence of increased workload as a barrier to further change.15 16 17 18 19 20 21

At the centre of the debate is whether general practitioners will carry out the work that arises from shifts in the balance of care without being given extra resources. In Primary Care: the Future, a precondition for taking a primary care led NHS further is to ensure that resources follow transfer of activities into general practice.10 Remuneration is also at the centre of the negotiations about the general practitioner contract. The profession's most recent document sets out a national definition of core services, which general practitioners are contractually required to do, as well as “non-core” tasks, which include many services with the potential to be transferred from secondary care and for which specific payment should be made.22 This, in effect, is the profession's version of a new contract.

Searching for evidence

What evidence is there that a primary care led NHS will necessarily increase general practitioners' workload and create the need for compensation? We undertook a review to identify what secondary care services are now provided in primary care, and to summarise the evidence available on the impact on workload.

The published literature was searched, using several databases including Medline, Social Science Citation Index, and BidsEmbase. The search proved difficult: keywords relating to the interface between secondary and primary care are poorly developed. Keywords relating to shifts in the balance of care (see box) were identified from general reviews. The term “workload” did not catch many studies relating to the shift, since data on the impact on general practitioner workload was frequently a minor part of a wider study and was not dealt with explicitly. For these reasons, a central part of the search consisted of consultation with experts and searches of the reference lists of selected texts. Studies were included only if the design made it possible to distinguish between the workload generated from shifts in the balance of care and the workload that would have been in primary care in any case–that is, controlled comparisons and “before and after” studies. In all, we surveyed more than 200 studies; the list is given in the full report, which is available from the National Primary Care Research and Development Centre.

Nature of the secondary-primary care shift

The 15 distinct areas in which a transfer of activities from secondary to primary care has been identified are listed in the box. “Shifted activities,” for which only the location of care has changed, can be distinguished from “substituted activities,” which also involve a shift in the type of healthcare professionals delivering the service. Workload implications for general practitioners arise only in the case of the substituted activities, but in many instances the additional work might be undertaken by other professionals within the primary care team.

Many general practitioners are now providing a range of services at their premises, rather than referring patients to hospital. Thus, most are likely to be directly involved in minor surgery and shared care for chronic disease management. Some general practitioners are also providing specialist diagnosis and treatment as an in house service. Those who have gained open access to, for example, surgical waiting lists, are providing preoperative assessment, and in a few places are providing postsurgical follow up of patients who are discharged to general practice rather than to outpatient clinics. In addition, the shifts to community based care and changed discharge practices in the acute trusts could have knock on effects for all general practitioners. Thus, general practitioners and other practice staff or community nurses are likely to be caring for patients discharged early from hospital, or after day surgery, as well as providing continuing care for severely mentally ill people, younger people with learning disabilities, and elderly people discharged from long stay and acute inpatient hospital departments. Developments such as general practitioners taking over routine work from accident and emergency departments and community based continuing care led by general practitioners are not, as yet, widespread. Consultant outreach and directly employed specialists are examples of shifted activities where the workload implications for general practitioners are minimal.

Evidence on implications for workload

Much of the literature on general practitioners' workload has arisen in the aftermath of the 1990 general practitioner contract and has focused on the impact on the contract's administrative requirements and increased participation in health promotion and prevention. Only 12 studies have provided evidence on the specific workload implications of secondary care services being shifted into primary care,23 24 25 26 27 28 29 30 31 32 33 34 and only one study has addressed the issue directly.23 This study examined the effect on workload of day surgery and the care required after discharge to the patient's home. The remaining 11 studies are all cost effectiveness studies that provide evidence which allows the evaluation of the effect on workload of four additional types of changes. Two randomised controlled trials evaluated the introduction of general practitioner led shared care for chronic disease,24 25 and one randomised controlled trial evaluated postoperative follow up in general practice instead of in outpatient clinics.26 Six studies27 28 29 30 31 32 gave information on the potential knock on effect of schemes in which well organised community based services substituted for hospital inpatient provision for mentally ill patients. Of these, four randomised controlled trials evaluated the introduction of comprehensive community based care, rather than standard hospital care, for acute psychiatric patients27 28 29 30 and two evaluated schemes for comprehensive community based care to replace long stay hospital provision for psychiatric patients.31 32 Finally, two studies indicated the resources needed for continuing community based care: one considered costs of general practice services for elderly patients with dementia discharged from long stay hospital to nursing homes,33 and another assessing a hospital at home scheme.34 Table 1 gives details of the information extracted from these studies.

The only scheme which clearly resulted in additional workload for general practitioners was discharge of patients with long term mental illness from psychiatric hospitals. Shared care schemes for chronic disease management for diabetes and asthma, and discharge of surgical patients to general practitioner follow up, had implications for workload, but these were minimal compared with the situation where no schemes exist, since patients visit their general practitioner regardless of whether they are attending outpatient departments. As predicted in the wake of the Audit Commission's publication of potential day surgery targets in 1992,35 the knock on effect of increases in day surgery on general practitioners' workload was minimal.36 Although the level of follow up required by day surgery patients varied between studies, this conclusion was reached by other studies as well.37 38

Figures from the workload studies have been used to model the impact of these changes on a typical week's work for general practitioners. Estimates were based on expected prevalence of the conditions in question in a typical patient list of 2000, and a mean consultation rate of three visits per patient per year.39 Implementing shared care for management of diabetes and asthma, long term mentally ill patients being discharged from hospital, and the Audit Commission's most optimistic day surgery targets would increase the total number of consultations each week by 5.6%: from 115 to 121.5 consultations. However, most of these consultations would be accounted for by shared care for asthma and diabetes, which does not represent extra workload. Patients were cared for in general practice before the 1990 general practitioner contract, and there would have been consultations with these patients whether or not there was a protocol for shared care. The effect of changes in day surgery and closure of long stay hospitals is an increase in the number of weekly consultations by only 0.8%–less than one extra consultation per general practitioner a week.

More evidence is still needed

The studies we reviewed cover only some of the changes that have happened in the 1990s, and the scope for making generalisations is limited. The studies evaluated schemes with well defined protocols and well resourced and organised facilities. This will not always be the case, and these are not the schemes that general practitioners are most worried about. Evidence is still lacking on the impact of early discharge from hospital, of day case surgery in which there is little regard to discharge arrangements, and of general practitioners taking over care of patients who otherwise might have been cared for in hospital–including visits to those in nursing and residential homes. Elderly people living in nursing homes require greater general practitioner input than does the remaining practice population over the age of 65,40 and disabled people living in purpose built housing units increase general practitioners' workload more than do other newly registered patients.41 These studies may indicate the workload implications of this type of work being transferred from hospital. However, studies are still needed which distinguish the work transferred from hospitals from that resulting from an aging population or patients moving house. Furthermore, in most studies, the definition of workload is confined to consultation rates, and other workload implications are not considered. Thus, very little is known about the extent of workload generated from the shifts in the balance of care.

Being unable to state whether a shift of services from secondary to primary care increases general practitioners' workload does not necessarily mean that there is not a problem. Rather, general practitioners have only anecdotal evidence to support their claims of appreciably increased workloads, and the NHS Executive has insufficient evidence to make informed decisions about remunerating general practitioners for the extra work resulting from the changes.

The present evidence does not point to significant increases in workload resulting from some of the most commonly implemented changes designed to shift the balance of care into general practice. More detailed analyses of the current concerns of general practitioners indicate that the secondary-primary care shift is not at the forefront of their thinking in relation to changed workload.42 43 Other pressures, such as increased expectations of patients and administrative burdens, are of greater concern. As the shifts gain momentum the picture may change, and there will be an increased need to identify mechanisms for ensuring that resources follow workload.

The problem of remuneration

One way of addressing the issues arising from the movement of services into general practice, and the feeling of general practitioners that they are overworked, is to find ways of regulating the way in which general practitioners are paid, rather than simply paying them more. Identifying a level of compensation nationally is made difficult by the inevitable local variations in workload resulting from differences in how widespread the various schemes are, and by variations in the number of relevant patients on practices' lists. This problem would be partly solved by agreeing payment rates for providing non-core services. However, to be evidence based, this approach still requires more research on the work required in connection with specific schemes. Furthermore, this way of remunerating general practitioners would not take into account the use that general practitioners already make of skills within the primary care team to relieve themselves of some of their work, as recommended in Primary Care: the Future.10 Tasks may be delegated to other professionals within the practice, who are already paid44, freeing general practitioners to undertake new activities without concomitant increases in workload. Furthermore, greater use of “fee for service” remuneration is widely believed to lead to demand on these services induced by doctors as a means of increasing income. There is a danger of increasing health service costs with little impact on the health of patients.45 Few studies have considered this assertion of perverse incentives empirically, and the results of those that have are questionable, due to their design and lack of controls.46 There is, however, strong evidence from Denmark that a change from capitation to mixed capitation and fee for service does lead to increased service intensity.47 The effects on use of resources and welfare of patients have not been examined empirically.45 46 48 More research is needed to evaluate whether the effects of such a change in remuneration systems are compatible with the goal of value for money and effectiveness in health policy.

Delegation

An alternative way of looking at the issue of services moving into general practice is to increase the emphasis on delegation of work to other professionals. Aside from preventive work and chronic disease management, delegation to nurses or the employment of new professionals to take on shifted tasks is in its infancy. Moreover, the degree to which general practitioners delegate work varies across Britain. This approach may not reduce health service costs, but it should reduce the impact of a primary care led NHS on general practitioners' workload. It remains to be seen whether other structural changes within primary care can be harnessed to facilitate this process. What is clear is that additional audit and research are required so that claims are based on evidence and not anecdote.

“The impact of the secondary-primary care shift on GP workload” is available from the National Primary Care Research and Development Centre, 5th Floor, Williamson Building, Oxford Road, Manchester M13 9PL.