Introduction
Throughout this study, both qualitative and quantitative data were collected, guided by a literature review, with the aim of developing a theoretical framework to inform future implementation and evaluation of alternatives to face-to-face consultations in general practice. This chapter describes the process of synthesising the findings from the various sources of data. This includes implications for future research, as well as the development of a web resource, based on our findings, to support practitioners and policy-makers who are thinking about the introduction of alternatives.
Methods for synthesis
A conceptual literature review informed our approach to conducting the focused ethnography in the case study sites. The analysis of the qualitative data followed a stepwise process, whereby the ethnographic team developed the coding frame that was applied to all of the data. As the data collection became complete, a summary document was produced by the ethnographic team for each of the codes, which was then shared and discussed with the wider team. Similarly, the analysis of the quantitative data (as described in Chapter 6) was also discussed within the team. As a team, we reached a consensus on the findings. It was then necessary to validate and optimise these further with a range of professionals, from a range of perspectives, at a stakeholder workshop, to ensure that any theory and resource development would be applicable to those for whom it was intended.
Stakeholder workshop
A range of stakeholders were invited to a workshop event on 18 October 2016 at a central location in London. The 25 attendees included GPs from participating case study practices (n = 5), policy-makers from NHS England with a particular interest in the introduction of digital technologies (n = 4), academics with relevant interests (n = 5), patient and public representatives (n = 4), one person from a primary care consortium and one individual from each of the following: the British Medical Association, the Royal College of General Practitioners, NHS Digital, NIHR and the Hurley Group (which promotes eConsult, an e-consultation software).
Following an introduction to the overall study and its aims and methods, the findings were presented through a series of brief presentations on key questions that we intended to use as the basis for the eventual web resource:
What were practices trying to do and why?
Whom is the alternative to the face-to-face consultation for and why?
How do we get implementation of alternatives to the face-to-face consultation right?
How will we know if implementation has worked?
Following each of the presentations, the stakeholders were invited to consider three questions within their table groups, which they then fed back to the room:
Do you think we have identified the right issues?
Is there anything we have missed?
What advice would you expect from our web resource?
The findings from these discussions were recorded on flip-charts and used to refine the structure and content of the web resource (described in Implications for practice and service delivery).
Patient and public involvement input to the synthesis
Patients, carers and members of the public inputted to the stakeholder workshop in two ways. First, four members of the public attended the workshop. These included people with experience of caring for an elderly parent, people with diabetes mellitus and people with depression. In order to ensure that young people’s perspectives were also included, we also invited members of Bristol Young Healthwatch, but none was able to attend. However, we had already elicited their opinions on alternatives to face-to-face consultations in a workshop specifically for young people, and were therefore still able to incorporate their views.
The public contributors who attended the stakeholder workshop actively contributed to the discussions, indicating where there may be areas of consensus and potential disagreement in the adoption and use of alternatives to face-to-face consultations between patients and health-care professionals. Specifically, they suggested that an aim to deliver consultations in an efficient manner, making best use of everyone’s time, would be something that would be shared by both patients and health-care professionals. They also felt that patients should be involved in any decisions about the method of consultation to be used, rather than this being a decision made solely by health-care professionals.
They thought that it was important that the web resource should include material that was aimed at patients and carers, to better inform them about the pros and cons of alternatives to face-to-face consultations. The PPI representatives also thought that the resource should include information on how practices can inform patients and carers on the introduction of alternatives to face-to-face consultations, recognising that one of our findings was that this was not always done well. They discussed issues such as whether to inform specific patient groups or to inform patients more generally.
The views of the PPI representatives concurred with our focus in this project. This included ensuring that the views of carers were considered, and that the use of alternatives to the face-to-face consultation was not limited to communication with GPs, but could include other members of the practice (e.g. nurses or nurse practitioners).
Programme theory
The underlying basis for this research programme was to gain a greater understanding of the use of alternatives by developing a programme theory which would set out how, under what conditions, for which patients and in what ways these alternatives may offer benefits to patients and practitioners in general practice. This would be based on a realist perspective to describe the provision of alternatives to face-to-face consultations in terms of:
In addition, we wanted to explore implementation in terms of moderating factors that influence the extent to which the process and outcomes are achieved, such as factors acting as barriers to, and facilitators of, practices offering alternatives to face-to-face consultations or to different groups of patients using them.
Context
The introduction of alternatives to the face-to-face consultation needs to be understood within the current policy context, which has consistently promoted the use of new technology to modernise health care.7,14 Over the last decade, the idea of online systems to improve access to health care has been repeatedly highlighted as an example of how the NHS will better respond to patients’ needs in the future.
However, our survey of practices demonstrated that, despite a great deal of enthusiastic policy-making and government incentives aimed at increasing the use of alternatives to the face-to-face consultation, with the exception of telephone consulting, these are not at all widespread. Even in the case study practices, where service providers have attempted to introduce innovation, our analysis of routine consultation data showed that use levels for alternatives to the face-to-face consultation were low.
The responses to our survey, and to our interviews, suggest considerable unease at the prospect of introducing new means of accessing a service that is already failing to cope with patient demand. Even though clinicians acknowledged that one motivation for introducing new technologies might be to control the perceived unmanageable workload, many felt that these technologies would increase, rather than reduce, demands on their time, or felt that they did not have the energy or time to put the required changes into place.
These reactions need to be understood within a local context in which general practice is under severe strain. There were many examples in our interview data of practice staff feeling unable to provide a functional service that met patients’ needs or policy-makers’ expectations, leading them to explore options for managing this demand, options that included alternatives to the face-to-face consultation. Interviewees, both patients and staff, described a system that they felt was falling short of providing reasonable access to primary care and this was a source of stress for both service users and providers.
In this context, for some practices, the rationale for introducing a face-to-face consultation was the need to ‘do something’ to change their practice organisation with respect to access to care for patients, even if they were not sure whether or not the change would lead to improvements. However, introducing a change to practice organisation where there is not the capacity to implement it effectively has the potential to lead to problems of implementation, and these were observed in our case study sites.
From our practice observations, a key factor in the introduction of alternatives in the case study practices in England was the financial support provided by the GP Access Fund.14 Many of the consortia of practices that were successful in bidding for funds included plans to introduce online or Skype consultations. This meant that practices within successful consortia were often given free access to e-consult systems (which would otherwise incur a cost), along with training and project management support. Two of our case study sites were trialling e-consultations funded this way, and one practice was using e-mail as part of a GP Access Fund14 project. One of the practices in Scotland was using an e-consult system as part of a free pilot offered by the e-consult company. The practice liked the system, but felt that it could not justify the cost and decided instead to incorporate features of the e-consult system into its own website, which would cost far less.
Mechanism and intended outcomes
In our focused ethnographic fieldwork, it was not always apparent that the reasons behind a decision to adopt alternative consulting methods had been clearly thought through and discussed within the practice. Members of the wider primary care team were not necessarily involved in the decision to implement, and, in the case of e-mail, were sometimes unaware of the fact that an alternative had been implemented. Decisions to implement were often led by one or two enthusiasts in the practice, with others going along with the decision and, on occasion, others openly disapproving.
Rationales for introducing new forms of consulting included more efficient management of demand, improving access to patients and to be seen as a modern practice. It was not always clear, however, that even when an expressed aim was vocalised, that there was a theoretical basis as to how the proposed innovation was likely to deliver that aim, particularly in respect of improved efficiency. There were a few exceptions to this; for example, in the remote and rural practice, video consulting was intended to overcome difficult, and sometimes impossible, journeys to the practice.
Interviews with patients suggested that they were interested in using these technologies (particularly telephone and e-mail), which they saw as a means of reducing the time they had to expend arranging to see and consult clinicians (particularly for what they termed ‘simple’ problems). The use of alternatives to the face-to-face consultation could also be much more convenient, as well as time-saving, particularly for people who have difficulty physically getting to a surgery, either for geographical reasons or because of illness. The asynchronous nature of some alternatives, such as e-mail, meant that patients could send a message at a time that suited them and then read the response later, rather than having their diary dictated by appointment availability. The use of e-consulting and telephone consulting was seen as a quicker form of access without needing to wait for an appointment, although this was not always the reality.
In summary, patients’ hopes were based around improved access, and there were practice staff who expressed similar views. However, in contrast, a clear motivation for practice staff was to manage demand, even if the potential impact on patient access to care was as yet unknown. There were examples in the interviews of staff hoping that alternatives to the face-to-face consultation would ‘keep patients away’ or act as a triage mechanism, in order to reserve face-to-face consultations for those patients who were felt to need them, but conversely, there were staff who hoped that patients would receive better access and more convenient care. Clinicians hoped that e-consulting systems would provide more structured information exchange, which would be more efficient. Both practice staff and patients believed that it might be possible to deal with simple problems more quickly and easily, so that longer face-to-face consultations could be used for problems that required proximity.
There was, therefore, a tension between two contrasting perspectives. One perspective was the aim to increase access, even at the price of increased workload (because, if there is currently a problem of poor access because of unmet demand, improved access would potentially lead to more patients consulting). The contrasting perspective was the aim to reduce practice workload, even at the price of placing restrictions on patient access. In this scenario, alternatives to face-to-face consultations were seen as a way of channelling patient demand in ways that would be quicker for the clinicians to deal with, even if this was not necessarily the patient’s preferred option. These contrasting perspectives highlight a number of important considerations and assumptions about how the introduction of alternatives to the face-to-face consultation might achieve the desired benefits.
First, one way in which these perspectives might be reconciled is if the use of alternatives to the face-to-face consultation leads to more efficient working, so that access could be increased, while also decreasing practice workload. Both clinicians and patients tended to assume that the use of alternatives would be time-saving overall, even if this was attributable to the redistribution of consultation time across, for example, different patient groups. However, there was a tendency for staff to ignore the workload occurring at the practice level. Doctors and nurses sometimes observed that they were expected to deal with alternatives to the face-to-face consultation in addition to, rather than instead of, their usual number of consultations, by slotting them in between face-to-face consultations or tagging them on at the end of the day. Staff in our practices expressed views suggesting that other members of the team did not fully recognise the impact on them. Reception staff were expected to add to their workload without any planning or allowance for this. There were few instances of practices taking steps to evaluate the overall workload associated with their use. This is an important point because, although previous studies have shown that telephone consultations are indeed shorter than face-to-face consultations (mean of 5.4 minutes vs. 9.2 minutes, respectively18), telephone consultations are also associated with a higher proportion of reconsultations. Thus, after taking account of the proportion of reconsultations telephone consultations do not appear to be time-saving overall.60 The fact that many practices implemented e-consultations only after they were given financial incentives to do so suggests that they were uncertain about the potential impact on workload, otherwise there would be a good business case to implement them without the need for any subsidy (unless the financial support was just to cover set-up costs).
Second, the ‘mechanism of action’ largely relates to implementation. Alternatives to the face-to-face consultation may improve access for patients if they are able to choose what form of consultation is most appropriate for them in a given situation. However, in some cases (related to the practice motivation to control workload rather than to improve access), practices used alternatives to the face-to-face consultation to control how patients gain access to care. For example, in some practices, receptionists suggested telephone consultations only once all face-to-face consultations were fully booked, or required patients to have a telephone consultation first. In some situations, this could be convenient for patients, but in other situations it was much less convenient for a patient if they had to ensure that they were available on the phone and to keep the whole day free because they did not know when they would be called.
On a related note, several practice interviewees described the need to ‘educate’ patients in using alternatives to the face-to-face consultation. This may simply refer to raising awareness of their availability, and we observed that some practices needed to put more effort into informing patients about the options that were available. However, it could also imply that staff felt that patients may use them in ways that the staff deemed inappropriate. This suggests that part of the mechanism of action was managing patients’ expectations of access to general practice.
Third, the argument that alternatives were useful for ‘simple’ problems assumed that both patients and clinicians can reliably tell in advance which problems are straightforward, and that both parties have a shared view about which problems can be managed remotely. Some of the interviews suggested that some staff do not have confidence that patients can always tell which type of consultation medium is appropriate. This relates to the comments about the need to educate patients, as described above, and the inconsistent way in which some practices asked patients to tell receptionists the reasons for some types of consultation request, but not others. This point is also supported by the finding that many clinicians applied alternatives to the face-to-face consultation selectively with patients that they deemed ‘sensible’, stating that they feared that these alternatives might be used inappropriately by other patients. This finding has also been observed in a previous study.19
Finally, there was an assumption evident across interviews with patients, doctors and receptionists that face-to-face consultations were viewed as the ‘ideal’ and that an alternative was second best, but better than no access at all. Some patients described not wanting to be ‘fobbed off’ with a telephone consultation, although clinicians sometimes expressed the view that alternatives to the face-to-face consultation were necessary because they could not meet demand for face-to-face consultations. Therefore, use was not necessarily a preferred option for either patient or doctor, but sometimes applied as a response to a problem of undercapacity; this was especially the case for telephone consultation. However, it was acknowledged that alternatives to the face-to-face consultation brought benefits that somewhat offset what was being lost in not having a face-to-face consultation, and for some patients the use of an alternative to the face-to-face consultation was clearly preferable.
Unintended consequences
Although patients (and some practice staff) saw improvements in access to care as a benefit, this did not take away from their worries about demand. Even if the use of alternatives to the face-to-face consultation does mean that consultations can be conducted more quickly and efficiently, there was a concern that increased access will increase demand, outstripping any efficiency gains. Concerns about supply-induced demand caused by offering access to alternatives to the face-to-face consultation were demonstrated by GPs; however, we did not find evidence about whether or not the provision of alternatives does indeed lead to supply-induced demand. Our case study practices experienced the opposite problem, in that use was so infrequent as to question the investment in these new approaches. It is unclear if low uptake results from GPs applying the use of alternatives selectively, patients not being aware of their availability or not wishing to use them, or problems with implementation.
A further unintended consequence expressed was the possibility that the provision of alternatives to the face-to-face consultation might lead to increased consultation rates among groups of patients who are confident with technology. These people may have fewer health needs, and self-care because access to primary care is currently difficult. Lowering the threshold to access would not only increase practice workload, but could also increase health inequalities. This point is considered further under the discussion about the impact on different patient groups.
Implementation
Contextual factors affecting implementation
A range of contextual factors affected which alternatives were implemented and how they were implemented, often in relation to practice population and geography. Our case studies illustrated extreme examples of this. The introduction of video consulting at the rural practice in Scotland was motivated by the difficulty of reaching a doctor by boat, whereas the use of e-mail consultations at another was partly related to the large student population. However, practice culture was also very important, with some of the case study practices introducing alternatives to the face-to-face consultation because they had a long history of being innovative, and they wanted to be forward-looking. As previously noted, in England, the support provided by the GP Access Fund14 was another contextual factor affecting implementation. Some practices had introduced e-consulting systems chosen by their local consortium, without necessarily having thought through how these systems would be used in their practice.
Moderating factors affecting implementation, barriers and facilitators
The GP Access Fund14 also acted as an important facilitator of implementation, providing an impetus to change, financial support and, in some cases, training and support with protocols for provision of a new service. Innovations introduced in this way were more likely to be accompanied by campaigns to raise patient awareness, including messages in the waiting room and in practice websites, although this did not necessarily relate to increased uptake levels.
It was clear that in several practices, introduction was driven by one or two ‘innovators’ in the practice. This could act as a facilitator of implementation, but could also lead to inconsistency. In some practices, it was evident that practice staff were working in different ways, not necessarily knowing how others were working, and with no formal practice policy on how to apply or use an alternative to the face-to-face consultation.
There were several barriers to implementation. These included a lack of training for practice staff, particularly receptionists. In some practices, receptionists felt that there was a lack of awareness from both patients and doctors of the complexity of their role. Furthermore, the use of alternatives to the face-to-face consultation had grown organically in most practices without a plan, and new ways of working had not necessarily been implemented in an efficient or organised way.
Structural factors also had a major influence on implementation. The use of video consulting in one practice was constrained by the inconvenience of using the equipment. Many practices experienced problems because of the limitations of their GP computer records systems, which made it difficult to include records of e-mails or e-consultations. Changes in the use of the telephone for consulting sometimes required more telephone lines than were available.
There were also more subtle factors that acted to impede or modify use, and these were related to the impact on professional identity. New ways of working could be perceived as a threat to aspects of work that were seen as fundamental to professional values. For example, a core tenet of general practice is the importance of the doctor–patient relationship, and forms of consulting that are not based on face-to-face contact may have an impact on that relationship. Alternatives to the face-to-face consultation have often been promoted on the basis of quick and convenient access to care, but clinicians in the case studies speculated that this may be seen as being to the detriment of continuity of care (associated with relationships), as well as potentially to the quality and safety of care.
Some forms of alternatives to the face-to-face consultation were associated with changes in professional roles, for example the greater use of nurses to conduct telephone consultations. This was seen positively by some staff, and has the potential to increase staffing capacity, but in other cases was perceived negatively, with some staff (both nurses and doctors) feeling that the nurse’s time might not be used appropriately. One example of how alternatives to the face-to-face consultation were associated with changes to professional roles is a system (not observed in our case studies, but widely reported during our research) in which practices contract with another organisation so that patients have the opportunity to have a telephone or video consultation with a doctor unconnected with the practice, and who does not have access to the patients’ records.18,154,155 Such developments challenge some of the key principles on which general practice is based, such as the provision of comprehensive, co-ordinated care provided through a single point of contact. However, this approach can improve access and expand the GP workforce by employing doctors working from home who cannot or do not wish to work in routine general practice. Given the scale of these changes, it is not surprising that the implementation of alternatives to the face-to-face consultation is met with resistance from some professional staff, which may have more to do with the implications for their professional values than it is related to the technology itself, and this was highlighted in the conceptual review (see Chapter 2).
Effects on different patient groups
One of the questions of interest for this research was the impact on different groups, including which groups might be most suitable for different forms and the possibility of increasing health inequalities if other groups are unsuitable. The key finding from our study is that, although individual patients’ perspectives varied, these could not be predicted based on factors such as age, sex or health status. In addition, there was a recognition that an individual patient’s reasons and ability to use a particular form of consultation may change over time and with experience.
Those who contributed to the stakeholder workshop felt that many of the assumptions about which patients benefit most were challenged by the evidence and, therefore, it would be more appropriate to consider how and when to use alternatives to the face-to-face consultation, rather than with which groups. Some forms have primarily been viewed as a way to respond to requests for quick access to care, but stakeholders felt that the same technologies might be more useful for follow-up and monitoring, rather than for initial assessment.
Although differences between patient groups were fewer than might have been anticipated, some concerns were expressed about the potential to increase health inequalities. This is supported to some extent by the findings reported in Chapter 6, that the characteristics of patients using e-consultation (more common in white, affluent, young adults) were in marked contrast to those of patients using telephone or face-to-face consultation. This could be interpreted as providing choice and improving access for people who find conventional forms of access to be inconvenient. However, it could represent inequity, by diverting resources and consultation time to the groups of patients with the fewest health needs. Given the very low number of patients using e-mail or e-consultations, it is important not to overinterpret these findings.
Implications for evaluation
Our aim was to understand how, under what conditions, for which patients, and in what ways, alternatives to face-to-face consultations may offer benefits to patients and practitioners in general practice and to determine which research methods were likely to be feasible to answer these questions.
The results of our research have provided useful data to inform the design of studies to undertake a formal evaluation of alternatives to face-to-face consultations. These include:
the number of practices currently using different types
the likelihood that practices and their patients will try these out
what outcomes are important to practices and their patients and what they expect to be achieved from introducing alternatives
to what extent alternatives are implemented differently in different settings and for which purposes
the feasibility of measuring the impact of alternatives on consultation rate and length, wider consulting behaviour and secondary care use
the groups of patients who are likely to be advantaged or disadvantaged by the introduction of alternatives
the types of methodology that are likely to be suitable to answer different research questions for different types.
The feasibility of answering these research questions depends on the level of development and current implementation of the technology under investigation.
One important consideration is the appropriate time at which to conduct an evaluation. Ideally, this would not be done too soon after implementation before the change has bedded down. However, neither should evaluation be conducted when it is too late to have an impact.
Potential research questions
In respect of alternatives to face-to-face consultations, our research has been partly about generating hypotheses, in order to create a logical framework for future definitive studies. Our work has indicated a need for further research on a range of questions:
Is it possible to improve the uptake of alternatives to face-to-face consultations and, if so, does this lead to benefits for patients and general practices?
What is the impact on access to care as perceived by patients (speed of access, convenience, timely access to care that meets their perceived needs)?
What is the impact on NHS workload and, in particular, on different sectors of the NHS (primary and secondary care) and different professional groups (doctors, nurses, receptionists and administrative staff)?
What is the impact on the quality and safety of patient management?
For which patients and for which conditions are different forms of alternatives most efficient and effective?
How do different forms of consultation change the content of the consultation?
How satisfied are clinicians and patients with different forms of alternative to the face-to-face consultation?
Our research has shown that telephone consultations are well established, whereas other forms of alternatives to the face-to-face consultation are very little used. Therefore, future research would need to apply different methods and explore different questions for these different technologies, and these are described below.
Telephone consulting
Telephone consulting is the only widespread alternative to the face-to-face consultation in common use. Although a large body of literature exists on telephone consulting [both qualitative and quantitative, including randomised controlled trials (RCTs)],60,156,157 most of this addresses the use of telephone triage of requests for same-day appointments and for out-of-hours work. There is some qualitative research on the use of telephone consultations for follow-up conditions and in long-term condition management, but little in the way of quantitative research, with a few exceptions for specific diseases.108,158–160 There is a need, therefore, to consider the impact on the quality and safety of patient management and NHS and patient resource use of the introduction of structured telephone follow-up of acute and long-term conditions.
The best way to determine quantitative outcomes (see ) would be to conduct a RCT. This could be randomised at the patient level, but would rely on GPs to recruit and randomise the patients, which is unlikely to be successful. Instead, a cluster randomised trial along the lines of the ESTEEM trial60 is likely to be more successful. An alternative might be a controlled before-and-after study. A qualitative process evaluation could be readily embedded in such a study, possibly within a realist evaluation framework. An economic evaluation with a wide scope would also be essential, because the impacts of this change in practice might be felt in several sectors of the NHS (e.g. role substitution in primary care, changes in the numbers of investigations or use of outpatient or emergency departments).
Proposed quantitative outcome measures for assessment of alternatives to the face-to-face consultation
In recent years, the use of telephone consultations as the main first point of contact for all consultations has been promoted and widely taken up by practices. We did not include such practices within our case studies, because these schemes are well established and now need to be evaluated using different research methods to assess outcomes.
E-consulting
We have identified very low uptake of e-consultations in this research. These systems incur a subscription charge, and, to be cost-effective, would have to be both widely used and reduce practice workload considerably. Research into these systems should begin with further investigation to determine why, as we observed, there is low uptake and how much time it takes a patient to complete the online forms and for clinicians to read them. If it appears that these systems can be made more attractive to patients, there is potential to see how their usage may affect the workload of the practice. Depending on the likely uptake of the intervention, this could be as an individually randomised trial (e.g. patients willing to use the system are randomised to those given immediate and delayed access after 1 year) or, if taken up by large numbers, as a cluster RCT. The study would explore the impact of the intervention on the outcomes described in . Such a trial should have an inbuilt qualitative process and economic evaluation.
E-mail consulting and webmail
In contrast to structured e-consultation methods, the use of e-mail for patients to enter into a two-way dialogue with the clinician is less well developed in the UK. It is a technology with which many are familiar and it is potentially inexpensive to introduce. However, as is clear from our case study practices, robust systems for handling e-mail have not been established in practices, and this would be difficult to achieve using standard e-mail (e.g. establishing with confidence the identity of patients using the system). Webmail has the potential to be more secure, requiring a logon from the patient. Such systems can be provided through practice websites. They would involve simple triage questions with respect to the nature of the problem to determine the suitability of the medium (e.g. speed of response, considered need for examination). Our work highlights serious concerns from clinicians about being overloaded, by lowering the bar to consultation and inappropriate use of e-mail as a means of fast-tracking to a face-to-face appointment. Unfortunately, there appears to be little in the way of quantitative research on the impact of the use of e-mail on workload, and on the analysis of the content of e-mails in comparison with face-to-face and telephone consulting for similar problems.
Clinicians would have to be reassured that the interventions will be introduced in a controlled manner. As with e-consulting, research should include Phase II research to develop and test the feasibility of robust webmail systems. Once such systems have been devised, then it is possible to consider introduction of webmail for a restricted number of interested patients who could be randomised to having immediate and delayed access. Embedded studies would explore how the content of e-mails compares with face-to-face and telephone consultations and patients’ and clinicians’ perceptions of the use of the medium. A reasonable intervention period (at least 1 year) would be required, as it takes time for people to become used to using new systems.
Video consulting
Although there is a relatively large body of literature exploring the potential for video consulting, many studies are small and descriptive.59,124 This is particularly true of primary care consultations, mainly demonstrating the potential for the medium, rather than exploring its impact on the content of the consultation or on GP workload or patient satisfaction. There have been no RCTs in general practice, and those studies in the UK have been confined to patient attitudes to video-consultation.
The major hurdle in undertaking research into video-consultation in the UK will be to encourage GPs and practice nurses to use it initially. Before undertaking any trials in the UK, it will be important to address the barriers to these consultations that we have identified. These include the attitudinal and logistical barriers identified by clinicians and the technical difficulties that may be encountered working within the heavily firewalled, low bandwidth systems of the NHS. Thereafter, it will be necessary to explore those types of patients for whom video-consultation is most appropriate, how video-consultations differ in length, content and quality from face-to-face and telephone consultations and patients’ and clinicians’ experience of these different media. At present, the use of video-consultation is so minimal that we are unable to answer these questions. Given the mismatch between the rhetoric and financial support being given to video-consultation compared with the almost non-existent use of it, the priority should be a cautious stepwise introduction, beginning with overcoming technical difficulties and establishing feasibility and potential impact, before widespread implementation of this approach. Once the use of video-consultation is more established, it may be possible to consider RCTs of the use of the medium versus face-to-face and telephone consultations.
Which patients
With the exception of telephone consulting, all new forms of consulting require an internet connection. Almost 8 in 10 households now have fixed broadband access at home and 61% of people regularly access the internet by phone or tablet.161 According to Ofcom in 2015, 68% of people aged > 55 years owned a smartphone, rising to 90% in younger age groups. At least until 2015, around 25% of non-smartphone users were converting to smartphone use each year.162 However, not all people who own a smartphone have large data allowances and this could be a problem for people with modest incomes who have no Wi-Fi connection. In contrast, home broadband usage appears to have peaked.161 People in the poorest households rely more on mobile phones than fixed broadband for internet access.162 It is clear that, for those consultation formats requiring internet connectivity, there may be a significant number of people excluded from their use, largely older and poorer people.
Analysis of our own data supported this finding and indicated that, although the demography of patients using telephone consulting was largely similar to the demography of patients attending the surgery, e-consultations tended to be carried out with younger and more affluent patients. There were insufficient instances of use of video consulting to determine the demography of users with confidence.
Interviews and observations with clinicians and patients also indicated that telephone consultations were more challenging for people who have communication difficulties (primarily those who did not use English as a first language or who had hearing problems), learning difficulties or cognitive impairment. However, for some of these groups, written communication such as webmail/e-mail and e-consulting systems might be helpful. Future developments would have to be carefully planned to ensure that the introduction of these technologies was not at the expense of those who may not have access to or be able to use them.
The choice of patient group relates closely to the question of how alternatives to the face-to-face consultation are implemented and for which purposes. If patients can choose from a range of consultation options that are equally available, then the patient group using each type of consultation will be determined by patients themselves. If use of alternatives to the face-to-face consultation is mandated by the practice as the default way to gain access to care, it will be important to facilitate other routes to care for these groups with particular needs, as described above.
Which technology
In summary, our research suggests that there are three types of technology that are priorities for robust evaluation, because they are being widely implemented in the absence of evidence. In each case, the research needed is less to do with the technology and more to do with the impact of the system redesign. The three priorities currently for evaluation of alternatives are as follows:
The use of the telephone for follow-up consultations; as previously noted, most attention has been given to the use of the telephone for initial consultations or triage, but there is great scope for use of the telephone for follow-up and monitoring of previously diagnosed problems, review of chronic conditions, support for self-management and simple queries. As we observed in the case studies, many practices use the telephone in a variety of ways that are often ad hoc. There have been several studies of telephone support for a range of specific chronic conditions,
158 but few studies of a whole-system redesign that maximises the use of telephone consulting for follow-up in an organised way.
159,160E-consulting systems are being heavily promoted, and their implementation has accelerated greatly as a result of the
GP Access Fund.
14 These systems often come with support and standardised protocols, and have often been acquired as a result of the decision of all the doctors in the practice. The use of these systems is quite formalised, which may facilitate research. Despite the attention given to these systems, there is very little evidence about their impact, and this is therefore a priority for research. However, in our case studies, we observed very poor uptake of these interventions, despite their being heavily promoted by practices. It would be important to explore the reasons for this, before embarking on a large-scale evaluation.
The use of ‘telephone-first’ models of access to care. This is currently being studied by another research team using mixed methods, including a before-and-after study of patient experience and hospital utilisation, qualitative interviews with patients and analysis of routine data about consultation rates, waiting times and continuity of care in a sample of practices. These studies will provide much-needed evidence, but will inevitably share the limitations of all such quasi-experimental studies, particularly the potential for confounding and selection bias. It is likely that practices that take up telephone-first models differ from other practices in important ways (particularly relating to local context and practice culture) that are hard to measure. Nevertheless, this is likely to be the best evidence available, as a true experimental study, such as a RCT, would probably be impossible to conduct.
Clearly, many of the issues above will pose a challenge to future research in the area. Particularly for those interventions that are little used currently (e.g. webmail/e-mail and video consulting), many clinicians will need to be persuaded of the possible long-term benefits, and will probably require some form of support/incentive to identify and agree the need for the change in their current practice.
Is there a need for a standard implementation of alternatives to the face-to-face consultation in undertaking research?
Our research shows that similar technologies are implemented differently in different practices. Some of these had been planned carefully and others had evolved, possibly not always with a great deal of thought. Implementations were sometimes somewhat piecemeal, with not all clinicians involved in their use and with no written or accepted protocols for handling them. It seems very likely that, regardless of the final implementation, one that has been planned and agreed with members of the practice team is more likely to be successful than one that has not. For the purposes of researching the impact of alternatives to face-to-face consultations, it makes sense to ensure that practices agreeing to take part in research should be supported to develop protocols for handling novel consultations in a way that suits their practice demography and existing management style to increase the chances of a successful implementation. However, it is likely, for example, that a solution that suits a remote and rural island practice will not necessarily suit an inner-city practice with a large non-English-speaking population. It is also possible that the implementation will evolve with time as practices gain experience. Alternatives to the face-to-face consultation are a good example of a complex intervention, and a range of research designs may be more or less applicable. A rigid, unalterable intervention, such as is normal for a traditional RCT, is unlikely to be appropriate. Regardless of approach, it will be important to specify the key features of the alternative to the face-to-face consultation (which, depending on study design, may be an intervention) that are fundamental to its intended purpose, and, therefore, must be implemented as planned, and those features that are flexible and appropriate for local modification.
Although there are many other potential uses of alternatives apart from the three listed above, these are less established, and attention should be given to further development and feasibility testing before definitive evaluation. In the case of some technologies, such as video consulting, for which there is little experience of use in primary care and different technological options are available, it will be important, prior to attempting to assess impact, to determine which are most accepted and used by clinicians and patients.
Measures of the process of care
This research has highlighted the range of factors that may act as mediators or moderators to the successful implementation of alternatives to the face-to-face consultation. Some of these factors are related to the presumed mechanisms and assumptions about how they might improve general practice (e.g. that managing an e-consultation will be quicker than a face-to-face consultation, leading to lower practice workload and shorter waits for a face-to-face appointment). In future research, it will therefore be important to assess these factors as measures of the process of care. This will help to illuminate how and why the intervention was effective, or if it was not effective, whether this was because of a failure to deliver the intervention as intended or a failure of the intervention to have the desired effect. Some factors, such as the number of consultations of different types, could be considered as measures of the process of care or as the outcome of the implementation. We have therefore considered both process and outcome measures together in .
Measures of outcome
The parameters best used to capture the outcomes of most importance will depend on the chosen methodology and the type of alternative to the face-to-face consultation under investigation. However, the key outcomes that matter to patients, practice staff and the NHS are listed in . Some of these outcomes are a direct result of the use of alternatives to the face-to-face consultation, whereas others may well be related and are important to patients (e.g. health-related quality of life), practices (e.g. staff morale) and the NHS (e.g. possible knock-on effects on emergency department attendance rates).
The parameters listed in reflect the priorities of patients and practitioners identified in this research and we have proposed suitable measures and made comments about their feasibility based on the experience of the research team.
Feasibility of practice recruitment
In order to evaluate alternatives to the face-to-face consultation in a rigorous research design, it will be necessary to recruit a sufficient number of practices in a range of settings. However, our survey of practices reported in Chapter 3 demonstrated a low level of willingness to implement. Therefore, a prospective study may be impossible to conduct, unless alternatives to the face-to-face consultation are developed that appear to offer advantages to general practices and are implemented for the purpose of evaluation. The only evaluative research that would appear to be possible at present would be studies of approaches that are being widely implemented, such as new forms of telephone consulting or the introduction of e-consultation.
Feasibility of patient recruitment
We did not seek to recruit patients on a large scale in this study, but previous experience suggests that it is possible to recruit at least 39% of people attending general practice to complete a questionnaire166 and 78% to allow access to their records for analysis of consultation rates.60 However, the feasibility of patient recruitment to research on some forms of alternatives to the face-to-face consultation will be constrained by the small number of patients using them.
Sample size
In the study reported in Chapter 6, up to 30% of consultations were conducted by telephone. With a telephone consulting rate of between 0.4 and 1.4 consultations per patient per year at different practices, it would be feasible to recruit a large number of patients within a short period. In contrast, e-consultations accounted for < 1% of all consultations in the three case study practices that provided them. Therefore, it would not currently be feasible to evaluate the impact of e-consultations unless uptake increases.
Feasibility of data collection for process measures and outcomes
above shows that most measures can be collected from patient-reported outcome measures or routine GP computerised records. Suitable questions are available from validated patient questionnaires for most of the relevant variables. Our research has demonstrated that routinely acquired data from GP computer systems can be interrogated to determine the number and types of consultations, including the number of subsequent consultations within 14 days. It was also possible to describe characteristics of the demography and medical history of patients using these consultations, as well as the professional characteristics of the clinicians consulted. Although not carried out in this study, it would also be feasible to collect data from GP computerised records of prescriptions and investigations.
However, the quality of recording of data in general practice is variable, particularly in respect of accurately recording consultation type. We found that telephone consulting was systematically under-recorded and that e-mail consultations were particularly poorly recorded. It was difficult to detect if a telephone consultation was for triage or a complete consultation (although the former may be inferred through searching for follow-up consultations within a defined period after the index consultation). Some forms of e-consulting systems (e-consult) were designed to remind GPs to record them, and as a result, were better recorded.
Although consultation type may not have been well coded, there were instances when this could be deduced from free text within the consultation record. Research utilising machine-learning techniques is promising in terms of identifying consultation type from free-text entry.167 Although we did not use these in our research, we are aware that algorithms have also been devised that can infer the likely consultation type from arrival time, start time and consultation length.168
The use of routinely recorded data in primary care therefore poses problems of accuracy and completeness. It seems unlikely that clinicians will improve the accuracy of their recording in the absence of changes to the design of current GP consulting software. Potentially, incentives may improve this. If this is not possible, it has implications for the size of studies that may be needed to demonstrate changes in outcomes.
The website resource
The stakeholder workshop was helpful in terms of crystallising some of the key messages from this study. Feedback from stakeholders was used to identify actionable messages for practices and policy-makers. We used these as the basis for a website resource, under the following headings:
Why do you want to introduce an alternative to face-to-face consultations?
Which alternative are you interested in?
Who is it for and why?
How do we get it right?
How will we know if it has worked?
The web resource was developed with the help of the website team at the University of Bristol (). It was agreed that the resource should be simple, accessible and intuitive, and designed to address the issues that the user would find most helpful. Given the variability in the rationale for, and implementation of, alternatives to the face-to-face consultation, the principle of the resource was to provide a self-appraisal and guidance tool, rather than providing standardised recommendations. The user would be guided to consider a series of key questions relating to the introduction of alternatives to the face-to-face consultation, and the responses to these questions would suggest things to consider and links to sources of guidance. In broad terms, the web resource was structured to follow the topics covered at the stakeholder workshop, that is, the need to consider the rationale for the introduction of an alternative, who the practice was expecting to use it and for what purposes, suggestions about pitfalls to avoid in implementation and ideas about how to monitor and assess the success of the initiative. Because we were aware that users might not access the website in a linear fashion, a side bar was present throughout the resource to allow the user to go back or progress onto another section. We were also aware that some users may not access the resource at the beginning if they had searched using specific terms, so the sidebar helped the user to navigate the site.
Screenshot of website resource. Reproduced with permission from the University of Bristol.
Once complete, the website resource was viewed and commented on further by the PPI representatives. The final website can be reached at www.bristol.ac.uk/primaryhealthcare/researchthemes/alt-con/resources/.
An outline of the structure of the resource is provided in Appendix 22.
The development of the website resource has been through a rigorous process, beginning with the ethnographic researchers and extended to the wider study team through to stakeholders and PPI representatives. The synthesis of the feedback from the stakeholder workshop with the study findings has been crucial. The study team feels confident that the website provides an informed and evidence-based resource to guide potential users of alternatives to face-to-face consultations.
