General Practice

Primary medical care outside normal working hours: review of published work

BMJ 1994; 308 doi: (Published 22 January 1994) Cite this as: BMJ 1994;308:249
  1. L Hallam
  1. Centre for Primary Care Research, Department of General Practice, University of Manchester, Rusholme Health Centre, Manchester M14 5NP.

    Fundamental changes in the delivery of primary medical care outside normal surgery hours are under consideration in Great Britain. Published research into the provision and utilisation of out of hours services shows long term trends towards decreasing personal commitment among general practitioners and rising demand from patients for primary and hospital accident and emergency department care. Wide variations exist regionally, locally, and between practices. Previous studies, however, have been limited in scope and provide an inadequate basis for assessing the potential impact of change. The overall demand for care across all sources of provision cannot be measured: there is a lack of data on costs, and evaluative studies comparing alternative patterns of service delivery have rarely been undertaken. A period of experimentation and evaluation of a range of options should precede the wider adoption of any particular models.

    Full responsibility for the primary medical care of registered patients whenever such care is needed is a key feature of British general practice. This feature, however, has often been questioned by general practitioners. The fact that most general practitioners would prefer to replace 24 hour responsibility for their patients with a more limited commitment is not a fundamental shift in attitude but the culmination of 25 years of decreasing personal commitment and increasing reliance on rotas and commercial deputising services. Over the same time patients have increased their use of hospital accident and emergency departments as an alternative source of care.

    Several innovative schemes - for example, cooperatives of general practitioners, walk in emergency care centres, and the employment of general practitioners within accident and emergency departments - have been adopted or proposed to reduce the strains in the current pattern of provision of primary medical care at night and at weekends. The Department of Health is prepared in principle to see the organised introduction of primary care emergency centres within existing constraints on expenditure. However, little information about such centres or the other options is available. Any proposals for change should be considered in the light of existing evidence from research. When evidence is lacking, caution and careful evaluation of other forms of service delivery should precede any major changes in patterns of provision.

    This paper summarises the published evidence to date on current arrangements to identify where further research will be necessary and to discuss the implications of the research for current debates about the future pattern of provision of out of hours care in general practice.

    General practitioners' out of hours arrangements

    Time spent on call is a shrinking component of general practitioners' workload. Between 1964 and 1977 the proportion of general practitioners on call five or more nights each week fell from 39% to 9%1. By 1991 < 4% were on call 16 or more nights during the week or every weekend over a period of four weeks.2 Time on call between 7 pm and 8 am weekdays and between 1 pm Saturday and 8 am Monday fell to an average of 26 hours a week in 1989-90, nearly four hours < in 1985-6.3 In a 1989 national survey nearly a third of general practitioners reported no regular personal commitment to their practice's night cover.4

    Wide regional variations in on call commitments largely reflect the availability of commercial deputising services.4 In the late 1960s these services began to expand to cover most large cities in Great Britain. In 1964 only 9% of general practitioners sometimes used such services.1 Between 1971 and 1976 the number of general practitioners subscribing to them doubled from 4000 to 80005 and the number of patients having contacts with deputies increased by nearly 50%6; by 1977, 42% of general practitioners were using deputising services (26% regularly).1 Studies suggest that the overall proportion of users has remained between 38% and 45% but that the proportion of regular users has risen.*RF 2-4,7,8* In 1989 deputies carried out 46% of night visits.9

    Although 45% of general practitioners do not have access to deputising services,2 changes in the structure of general practice have facilitated the growth of practice rotas, an alternative means of reducing personal commitment. The proportion of general practitioners working single handedly fell from 43% in 1952 to 11% in 1989; the proportion in groups of four or more rose from 9% to 55%.10,11 Two thirds of general practitioners now use a practice rota for some or all duty periods, and up to 30% use a rota in collaboration with other practices at least some of the time.2,4 The case for larger, cooperative rotas in which several practices provide a service in an area was argued as early as 1984,12 and this movement has recently gathered momentum.13

    Few criticisms have been levelled at practice rota systems. Suggestions that lack of personal knowledge of individual callers may present a problem for general practitioners on a rota are not supported in published works.*RF 14-18* General practitioners believe that rotas are an effective means of balancing their own and their patients' needs.12,19,20 Patients' satisfaction is reportedly high.21,22 No systematic evaluation of cooperatives has so far been undertaken.

    Commercial deputising services have provoked considerably greater controversy. It has been argued both that they provide essential relief for general practitioners and that they undermine the principle of continuing personal care.*RF 20,23-30* Although general practitioners are satisfied with the standard of particular services,31 questions have been raised about deputies' qualifications, the quality of care provided, and the impact of such services on demand.*RF 32-37* Visits by deputies have been linked to significantly lower patient satisfaction.*RF 38-40* Guidelines on restricting the use and controlling the standard of commercial services apparently have not been consistently applied.41,42 but little research into deputising services has been conducted recently.

    Despite a reduction in the volume of out of hours care provided by individual general practitioners this aspect of their workload contributes disproportionately to feelings of stress, fatigue, and fear of violence.*RF 25,26,43-48* In an extensive recent survey 57% of general practitioners believed that 24 hour responsibility was outdated; 82% agreed that it should be possible to opt out and 73% said that they would like to do so.2 Around two thirds supported limiting their commitment to between 8 am and 7 pm and nearly a quarter would not be prepared to participate in any alternative arrangement for providing out of hours care. Support for ending 24 hour responsibility has come from both the General Medical Services Committee and the annual conference of local medical committees.49,50

    None the less, some studies of general practitioners show a professional commitment to providing care at all times and a disquiet with the perceived consequences of becoming a 9 to 5 service.*RF 20,28,29,51-53* The Royal College of General Practitioners has questioned the very term out of hours, asking what a doctor's hours are.43

    Demand for out of hours care

    It has never been possible to assess the demand for primary care outside normal surgery hours nationally. Evening and weekend day calls to general practitioners are not routinely recorded, and studies are generally restricted to individual practices or small areas. Between 1987 and 1989 five such studies reported annual rates of contact outside normal working hours to be between 130 and 176 per 1000 patients.*RF 15,19,54-56*

    The introduction of a fee for night visiting in 1967 enabled night visiting rates to be compared nationally. Between 1967 and 1976 rates rose from 4.3 visits per 1000 population to 10.1 per 100035 By 1981 a major study of calls in one health district reported a visiting rate of 15.5 per 1000.18 Wide variations between individual practices and practitioners contributed to these averages.*RF 17,51,57-61* Between 1985 and 1989 four studies of single practices or small areas reported rates between 14 per 1000 and 35 per 1000,15,19,54,62 suggesting that demand has continued to rise.

    In some areas general practitioners are no longer the primary source of care out of hours. In addition to the role played by deputies, hospital accident and emergency departments are said to handle up to 47% of first contacts.63 Unfortunately, studies in accident and emergency departments rarely discuss arrival times*RF 64-70*; if they do they are imprecise in defining periods outside general practitioners' surgery hours71 or are unhelpful in reporting attendance rates at those times.1,72 None the less new attendance at accident and emergency departments at all times of the day and night have risen sharply and continue to rise, from 105 per 1000 population in 1961 to 212 in 1983 and 241 in 1991.73,74

    Measuring the demand for out of hours care is further complicated by the poorly documented role of cottage or community hospitals run by general practitioners as an additional source of emergency services.*RF 75-82*

    Need for out of hours care

    Many studies describe how patients perceive and interpret symptoms and what factors trigger a decision to seek medical care. Such studies, however, do not relate specifically to out of hours consultations, and it is beyond the scope of this paper to consider them in depth.

    Some general practitioners think that few patients call them outside surgery hours with problems the patient knows to be trivial.51,58,60,61,83 A large proportion of calls, however, are subsequently judged by the doctor to be medically unnecessary. The terms used to describe these calls imply varying degrees of misuse of services - for example, irresponsible,51 questionable,82 completely unnecessary,17 and could have waited until the next working day.57 Night calls are much more likely to be judged genuine emergencies or at least reasonable. In various studies between 41% and 60% of all out of hours calls have been classified in categories denoting that they were unnecessary compared with between zero and 8% of night calls.17,51,57,58,60,61,82 Visits to children are less likely to be judged urgent than visits to elderly patients,38,84 and young children are heavily overrepresented in out of hours calls to general practitioners and their deputies. Children under 5 years old constitute between 6% and 7% of study populations and are generally said to be the reason for between 22% and 30% of all calls.*RF 14-17,20,32,36,37,55,57,82,85,86*

    * General Practitioners may not be the primay source of care out of hours

    (fig omitted)

    Social deprivation may lead to high contact rates*RF 19,59,87-90*; in addition, high expectations in areas of comparative affluence have the same effect.1,91

    Many studies have tried to define and measure inappropriate attendances and attendances for primary care reasons at accident and emergency departments. Recently, the proportions of patients whose attendance falls into these two categories have been estimated at between 29% and 70% of all new attenders.*RF 64,66,67,70,92-94* No standard criteria have been used in making these judgments.

    Though some of these patients will not have needed any professional medical care whatsoever, the rise in attendances at accident and emergency departments has been ascribed, in part, to a failure by general practice to provide sufficiently comprehensive and accessible primary care services.64,71,92,95 Yet studies show that only 3-6% of patients will have attempted to contact a general practitioner before attendance.64,66,69,71,93,96 They are reluctant to bother general practitioners, fearing it may damage the doctor-patient relationship.39,93,97 Some anticipate long delays in receiving help from a general practitioner or deputy whereas accident and emergency departments are immediately available.*RF 64-67,69,72,93,96,98*

    Patients also fail to understand the role of accident and emergency departments,*RF 64-66,93,96,99* and the reasons why they attend are complex. Although all trauma is commonly believed to be the province of accident and emergency departments rather than general practitioners,65,67,72,92,97 factors such as geographical proximity, where the need for care is experienced - for example, away from home - the role of authority figures, anticipation of referral by a general practitioner, and perceived severity of symptoms are as much a part of the decision making process as access to, and confidence in, general practitioners,*RF 65-69,70,72,92,96,97,99-101*

    Few authors have suggested ways in which primary care needs can be met effectively for patients who choose to bring them to hospital, though there have been some moves towards this.*RF 67,95,102-104* An innovative scheme with triage by nurses and a primary care stream staffed by general practitioners within an accident and emergency department is being evaluated.67,102,105 However, another recent attempt to organise an out of hours primary care emergency centre adjacent to a hospital accident and emergency department met hostility from hospital management.106 Although studies dealing with attitudes of accident and emergency staff towards their workload are rare,99,100,107,108 attenders who should have seen their general practitioner have been seen as being at the root of many problems and their re-education has been viewed by many as a priority.*RF 64-66,68,109-110* This is part of a complex picture in which prestige among hospital doctors is related to distance from the undifferentiated mass of patients. Staff in accident and emergency departments are not shielded by referral from general practitioners, which restricts the access of patients to other specialists. Low prestige has caused, and been reinforced by, a lack of senior appointments, particularly of consultants; by poor facilities; and by inadequate funding.

    Future of out of hours primary care services

    Many questions remain unanswered about the demand for, and the provision and utilisation of, primary care services outside surgery hours. Studies have rarely addressed more than one aspect of the subject and are often limited to a small geographical area, a single practice, or one hospital's catchment area. As a result published reports show a confused and confusing picture. The dearth of major studies examining utilisation across all sources of provision is particularly problematic. The relative contributions made by general practitioners, deputising services, accident and emergency departments, and cottage hospitals are impossible to assess nationally. The true demand for out of hours care is also impossible to assess. Because of this the likely impact of any changes in the way in which services are delivered cannot be assessed.

    Opposition by general practitioners to the 24 hour contract is unlikely to be solely the result of the amount of time spent on call. By increased use of commercial deputising services and rotas within and among practices most general practitioners have considerably reduced the on call component of their workload. Instead, 24 hour responsibility seems to be unacceptable because of the increased workload and stress during daytime. Thus, changes in the method of delivering out of hours care that further reduce the hours worked on call but do nothing to address the problems of general stress and dissatisfaction may be only partially successful in persuading general practitioners to retain 24 hour responsibility. The Department of Health has indicated that it is not prepared to consider removing this contractual obligation.69 Reports that its position might be changing have subsequently proved unfounded,111,112 though changes to the rules to release general practitioners from responsibility for the acts and omissions of properly registered deputies will be made.113

    Among the suggestions mooted for alternative arrangements, the formation of cooperatives providing services within a defined geographical area is finding growing acceptance among general practitioners.13,114 As I have pointed out, however, there are no published evaluations of this type of service. Different sizes of cooperatives and methods of funding and organisation have not been compared. Whether patients are able to distinguish between a visiting deputy and a general practitioner from a cooperative has not been established. It has been suggested, though not proved, that use of deputising services leads to increased demand. If this is so and patients make no distinction between deputies and general practitioners from cooperatives, cooperatives may also lead to higher demand.

    Furthermore, areas most suited to cooperative cover often already contain commercial deputising services. Commercial operations may be jeopardised if they lose a considerable number of subscribers. No major studies of the work of deputising services have been conducted recently, and it is by no means clear what effect the 1984 guidelines on use and control have had. In districts where extensive liaison and monitoring efforts have aimed to ensure a high quality of service general practitioners may be reluctant to exchange this quality for the unknown and unmonitored standard of care provided by a cooperative. With the demise of a deputising service, subscribers who cannot or do not wish to join a cooperative may find themselves without any relief.

    A more radical innovation, walk in or book in primary care emergency clinics, is being explored by the GMSC and the Department of Health.113 Patients able to travel would be expected to attend these centres. Greater flexibility in the use of funds for general medical services would be allowed to encourage this development. The effect of centralised services on demand cannot, however, be predicted. Home visiting is expected as an integral part of the primary health care system, and the proportion of patients prepared to travel to such emergency clinic facilities while a choice is available is anecdotally reported to be low (though this is difficult to reconcile with patients' increasing propensity to attend accident and emergency departments).

    Alternative systems of care that restrict home visiting may reduce demand among some sectors of the population, though the effects of this may not be wholly desirable. Detrimental delays in seeking care may result. Requests for daytime home visits may rise or some patients might come to view an emergency clinic as their normal source of care. Unless permanently open and staffed, clinics sited adjacent to or within an accident and emergency department might well increase the primary care workload of the staff of the department.

    Any dedicated, centrally provided service employing permanent or sessional staff, or both, would be unlikely to be as cost effective as the current system, but the true cost of providing primary care services out of hours is not known. No research has been undertaken on this subject and several features of the current system would make it difficult to conduct. With the current need to consider and compare alternative methods of providing an out of hours service, lack of data on cost is a fundamental problem.

    An alternative system which could be implemented nationally is difficult to envisage. Each of the main options proposed merits further exploration, but each is predominantly suited to urban areas, where deputising services and opportunities for extended rotas already exist. Patterns of service need to be considered for rural areas with low population density and limited access to support services, which rely heavily on the commitment of individual general practitioners.

    The demand for out of hours care also needs to be considered. Plans to extend the Help Us to Help You campaign to educate patients would benefit from examining possible relations between demand and daytime surgery hours, waiting times for appointments, and time allowed for appointments. If more appropriate use of out of hours services is to be encouraged these factors and patients' perceptions of what constitutes a medical emergency will require further investigation.


    In summary, the lack of comprehensive information on the current provision, utilisation, demand, and cost of out of hours primary care services and the limited attempts to evaluate any alternative system argue the need for an experimental phase, in which new approaches to service delivery are explored and evaluated, before the widespread adoption of any particular option.

    The views expressed in this paper are my own and do not necessarily represent the views of the Department of Health, which funded the work on which it is based.


    View Abstract