Remote care in UK general practice: baseline data on 11 case studies

Overview of participating practices

In Table 2, we provide a one-paragraph summary of each practice, which we have anonymised. We have classified each practice by index of multiple deprivation (from 1 = most deprived to 10 = least deprived decile) and by digital maturity using the following five-point scale²: • (traditional – few or no digital innovations or strategy), •• (traditional with lone innovator – one person keen and attempting to introduce digital innovations and services), ••• (digitally curious – experimenting with digital innovations but not planning or implementing these strategically), •••• (digitally strategic – investing in digital innovations and services, and in some cases strategically disinvesting in them) and ••••• (system-oriented – confidently providing a range of digital services and seeking to support others to do the same).

As shown in Table 2, the 11 general practices have a wide geographical spread covering inner-city locations in Scotland (Range Park, River Road), remote towns and villages in Wales (Carleon, Rhian), and various settings in England including major cities (Towerhill, Westerly), smaller cities and towns (Newbrey, Camp St, Ogden East, Queens Road) and villages (Fernleigh). Whilst we achieved wide demographic and geographical variation across three jurisdictions, our sample did not include any practices from the north of England or Northern Ireland.

Socio-economic status of the populations served ranges from very deprived rural (Carleon) and urban (Ogden East, River Road, Range Park, Westerly) to fairly affluent rural (Fernleigh) and urban (Towerhill). We deliberately oversampled from deprived localities—for example, whilst one practice (Fernleigh) is in the top decile for Index of Multiple Deprivation, three (Rhian, River Road and Ogden East) are in the bottom decile. Some practices (Camp St, Newbrey, Queens Road) have a very mixed population with some ‘postcode pockets’ of deprivation. The ethnicity of practice populations varies from 35% White with an extremely diverse ethnic mix (Westerly, Queens Road) to 99% White (Carleon); both Welsh practices (Carleon and Rhian) have a high proportion of Welsh-speaking patients and staff. The inner-city practices in Scotland and England (Range Park, River Road, Westerly) have young populations and quite high list turnover; Fernleigh (serving affluent retirement villages) has a more stable population but a high proportion of elderly.

Total practice list sizes vary widely from 31,000 (Camp St) to 2300 (Range Park). List sizes per full-time and salaried partner range from 2300 (Rhian) to fewer than 1000 (Range Park). In terms of staff mix, numbers of nurse practitioners, advanced health care practitioners, nurses and other health care support staff differ widely across practices. Many teams in the English practices are highly multidisciplinary, allowing a high degree of functional flexibility and providing the GPs to focus more on complex cases or specialty interests. The smaller Scottish and the Welsh practices have notably fewer non-medical staff.

Exact numbers of administrative and support staff are hard to capture (in some cases, such data are unavailable; in others, they change frequently). It is clear that some practices have a sophisticated division of labour among reception and support staff as well as among clinicians, and also some well-defined data management and IT support roles (Camp St, Fernleigh, Towerhill and Westerly appear advanced in this regard). In some other practices, the support roles for digital services do not exist at all or are less well developed.

Many practices are involved in non-core activities. Eight of the practices (Carleon, Camp St, Fernleigh, Ogden East, River Road, Rhian, Towerhill, Westerly, Queens Road) are teaching and training practices and two (Fernleigh, Towerhill) were involved in research before joining this study. Some practice members undertake additional activities, for example Carleon clinicians cover a community hospital; Fernleigh is a dispensing practice, and Towerhill partners are involved in local medical politics, organisation and management, with one partner working on the GP Federation board. One Range Park GP is a longstanding member of the Local Medical Committee. These activities suggest that whilst our sample is diverse in many dimensions, most or all are more outward-looking than average.

All practices have many core values in common. Interviewees in every practice, for example, talked of the practice’s commitment to its local population; a desire to provide high-quality, evidence-based, patient-centred care (and in many cases, providing high-quality care was seen as implicit rather than articulated as a value, though safety was occasionally mentioned); to be a happy and cohesive practice team with attention to staff training and wellbeing; to use multidisciplinary and holistic approaches in order to address illness in its social and cultural context; and to minimise inequalities of access and provision. Whilst interviewees in all practices said that continuity of care was valued, many described a trade-off between continuity and (for example) efficiency or timeliness (see ‘cross-cutting themes’ below).

Innovation and digital maturity

Interviewees from almost all participating practices depicted their organisation as forward-looking and keen to innovate, using terms like “dynamic and positive”, “forward thinking” and “aiming to embrace change”. These comments may reflect the fact that any practice which is prepared to join a research study on digital innovation is to some extent keen to innovate.

Our familiarisation process suggested, however, that practices varied widely in their organisational antecedents for innovation as outlined in a systematic review of the diffusion of innovations literature²¹. This showed that organisations that are able to introduce innovations (should they judge them appropriate) are distinguished by a number of features: structural preconditions (larger size, a flat management structure, devolved decision-making, a clear division of labour and well-differentiated roles, and slack resources that can be [re]deployed), absorptive capacity for new knowledge (a high level of pre-existing knowledge in relevant areas, the skills and systems to capture and distribute knowledge within the organisation, and wide internal and external networks), and receptive context for change (leadership, strategic vision, clear goals, a climate where it is acceptable to take risks, and high-quality data systems to monitor progress). But even when an organisation is able to innovate in general, it may assess a potential innovation as a poor fit with its focus and mission (see ‘readiness’ below).

Some practices in our sample (notably, Camp St, Fernleigh, Westerly and Towerhill) appear to have many of the structural and cultural preconditions for innovation. They are large in size, have some slack resources (at least compared to our other practices), and have leadership, strategic vision and clear goals for developing further. They also have high absorptive capacity for new knowledge (with technology-savvy staff, high-quality infrastructure and rich internal and external networks). Some practices possess certain preconditions for innovation but appear to be held back by insufficient slack, heavy workload and (in some cases) outdated premises and systems. As an informant in Newbrey put it “overwhelming workload … leaves no time for innovation.”

This background helps explain where practices currently lie on our digital maturity scale, which reflects three dimensions of maturity⁷: readiness (strategic alignment, leadership, resources), capability (what is currently installed and possible) and infrastructure (the technological and human infrastructure to support digital innovation).

The Remote by Default 2 practices illustrate the full range of digital readiness. At one end of the continuum, Towerhill has a clear and bold vision to embrace digital innovation and views digital access and consultations as an excellent fit with its strategic goals. At the other end, practices such as Range Park have made a clear strategic decision not to prioritise digitally advanced forms of remote care because they do not feel it is currently right for their patients (and perhaps also for their staff). Other practices present a more mixed picture. Newbrey, Queens Road, River Road and Westerly, for example, whose practice populations are socio-economically and ethnically mixed, are progressing cautiously with specific digital innovations while attending carefully to patients who are digitally disadvantaged in various ways.

In terms of digital capability, the practices also represent a wide range (though all have some digital capability). All practices are using telephone access and telephone consultations in much greater numbers than before the pandemic, but their telephone systems vary significantly in their dependability and functionality. One limitation is the sheer number of calls the system can cope with. The existing telephone systems in several practices (Camp St, Newbrey, Queens Road, Range Park) are described by staff as at or above capacity, and patients in even more practices describe not being able to get through on the phone. One practice (Towerhill) has recently introduced a sophisticated cloud-based telephony and triage system which will offer greater capacity.

Most practices have the capability to provide online consultations—most commonly using the accuRx add-on for SystmOne. A few have plans to adopt more advanced systems. As with phone, online consultation systems can be more or less sophisticated and have more or less capacity—a technical issue which we address in the cross-cutting theme below.

Many of the practices developed remote systems rapidly in response to COVID-19 and some are pulling back from these, partly because infection control restrictions around in-person consultations have eased. In some practices the changed working patterns inflicted by COVID-19 were very gratefully given up; GPs and nurses had described days spent only on the telephone (“call centre medicine”) as cognitively challenging and unfulfilling. Most practices appear to have continued only with the digital services that appear to add value outside the pandemic context.

A good example of this is video consultations. Only one practice (Range Park) never used video; the remainder had the capability, but many did not persist with it. A few practices (Ogden East, Towerhill, Westerly) are still routinely using video consultations and now have established systems and protocols for targeting them appropriately. Four practices (Carleon, Camp St, Fernleigh and Rhian) initially introduced but then abandoned video, and two more practices (Newbrey, River Road) now use this modality very sparingly (“once in a blue moon” as one previously enthusiastic GP said). In a previous study conducted in the first 18 months of the pandemic, we wrote about the widespread non-adoption and abandonment of the video option in UK general practice⁸.

The quality of infrastructure to support digital services also varies hugely across our sample of practices. At the more advanced end, Towerhill, Fernleigh and Westerly upgraded their infrastructure before the pandemic and were in a strong position to support a move towards more digital services. At the other, Rhian and Range Park are struggling with the most basic infrastructure and cannot make progress (even if they wished to) until this is improved. Scotland had a country-wide initiative to develop the infrastructure for video consultations before the pandemic⁵, though this was to a large extent separate from general IT infrastructure development.

Each practice’s digital maturity is thus a combination of its strategic readiness, its existing capability and its infrastructure. Table 3 shows examples of each level.

In the next section, we describe some themes emerging from our 11 practices with their varied histories, characteristics and patient populations. These include access, triage and capacity; other aspects of quality and safety of care; meeting the needs of the disadvantaged and digitally excluded; staff wellbeing and training; and technologies and their associated infrastructure.

Cross-cutting themes

THEME 1: Access, triage and capacity. Many GPs in our sample expressed concern about whether they were “seeing the right populations”. Whilst they would escalate a situation if they believed there was a clinical need (for example, bringing someone in to be seen in person), there was a concern that they may not always become aware of such need. However, demand is high and rising, and staff in many practices used flooding metaphors (“swamped”, “deluged”, “opening the floodgates”) to describe the volume of telephone calls and digitally-enabled requests for consultations they were having to deal with. This sense of losing (or having already lost) control of the threshold to primary care was palpable in many interviews. Hence, in the current context, we consider access as playing out in tension with capacity.

Every practice sought to avoid the problem of a patient in need being unable to access care. Some practices serving non digitally confident populations (Carleon, River Road, Rhian, Range Park, Westerly) have until recently relied mainly on telephone or walk-up requests that are triaged by reception staff. However, telephone queues—and physical queues at the front door—are becoming unmanageable in these practices. Partly for resource reasons, one or two practices appear to be stuck in a situation where clinicians and support staff are simply working harder (and becoming burnt out) as demand inexorably rises; others have introduced changes (Westerly, and more recently River Road, introduced online consultations; Rhian have hired a new business manager whose brief includes rationalising the appointments system).

Ogden East, also serving a deprived population, uses a different system—no triage by receptionists but a GP-led call-back system to every patient. One GP commented that this system allowed them to practice better care: they felt more in control of the day, reserving more time for more complex patients. They reflected that in such calls, they handled the query differently to how they might have done in an in-person appointment—they could be more candid and (for example) suggest to a patient that they go away and look something up. Ogden East also offer appointments on request to any walk-in patients.

Our baseline data reveals a tendency for online consultations to be used to take the pressure off the front desk. In Ogden East, for example, staff encourage patients to fill out an online consultation request if they are unhappy with the length of time they need to wait for an appointment. Yet several practices (e.g. Camp St, Fernleigh, Newbrey, Ogden East, Rhian) have found that online consultations are inefficient (because they collect irrelevant data, and because some patients use them for problems that staff consider inappropriate, especially if the system is available at night), potentially unsafe (patients have been known to use them for chest pain) and stressful (because of the sheer numbers coming in). In some practices (e.g. Rhian) the capability for online consultations is present but few patients use them (probably because the more traditional telephone and walk-in booking system is accessible and preferred by staff and patients). Online consultations appear to increase clinician workload while relieving workload for support staff (in another flooding metaphor, one GP described having to “wade through” multiple online consultation forms).

To control the digital “floodgates”, several practices (Fernleigh, Towerhill, River Road) now restrict the time slots where online consultation forms are available, and other practices are considering doing this. This move was described by one interviewee as a shift from “demand-driven” to “capacity-driven” provision. It is worth noting that in the practices where such digital gatekeeping is heavily used, support staff shared stories of near-miss cases where patients (typically elderly and with multiple health problems) either gave up trying to contact the practice or attended Accident & Emergency. In response to such critical events, one practice (Fernleigh) plans to move to clinician total triage.

One practice (Camp St) has introduced non-digital triage tools (standard operating procedures, guidance) and trained particular groups of staff in using these. Camp St also operates a ‘safety valve’ system where extra appointments are added to the system once it gets full (usually about 10 am every morning); this practice also has a safety valve for when the duty doctor is overrun, whereby other GPs and advanced clinical practitioners are asked to take up some of the load although this was noted as sometimes causing tension between GPs. Others (e.g. Fernleigh) make selected use of email to communicate with patients and signpost them to the appropriate part of the system.

In Westerly, the reception team have been trained to undertake an initial assessment of appointment requests, advising patients with minor illnesses to contact a pharmacy and booking some patients into appointment slots according to the practices booking rules (e.g. smear requests booked into a nurse in-person appointment; asthma reviews booked into a nurse telephone appointment). There is also practice guidance about booking in-person appointments for patients with selected symptoms (such as abdominal pain). All other appointment requests – whether made by phone, in person or online are allocated a full appointment slot. Short ‘triage’ slots to quickly review an online consultation have not been introduced.

THEME 2: Other aspects of quality and safety of care. We were surprised that many aspects of quality and safety of care, including how to achieve timely diagnosis of urgent or serious illness, how best to manage long-term conditions, and how to deliver care for patients with complex needs, were largely absent from our dataset of initial interviews. National policymakers with a safety brief knew of rare examples of ‘never-events’ with possible links to lack of a face-to-face assessment (e.g. death of previously healthy young adult from operable acute abdominal condition), but they emphasised that formal audits of telephone assessments (in which an experienced clinician reviewed an audiotape of the call along with the written record) identified the vast majority as high-quality and safe. It appears that clinicians undertaking telephone assessments have a low threshold for arranging an in-person assessment if indicated—hence the use of telephone by default tends to be inefficient rather than unsafe.

Practice staff seemed to take quality and safety of care as given, so long as patients could be seen in an appropriate and timely way. For this reason, access and triage (see above) were depicted as mission-critical to quality and safety. The only other aspects of quality and safety which came up repeatedly in our baseline interviews were continuity of care and risks of technology failure.

Continuity was universally depicted as an aspect of high-quality care. However, our interviewees held different views on what continuity was (one-to-one continuity of care, continuity within a small sub-team, or continuity of information), how continuity should be delivered in practice, and the trade-offs against other practice priorities. No practice in our sample had a strict personal list system, but some (e.g. Camp St, Fernleigh) had a ‘usual GP’ arrangement and one (Westerly) operated a ‘buddy group’ system in which GPs and advanced clinical practitioners were clustered in small groups so patients were highly likely to be allocated one of a small number of known clinicians. Queens Road offers patient choice to see a named GP through its online bookable system but encourages patients to see GPs with expertise in specific clinical areas across other practice sites in its medical group. One practice (Towerhill), which has a mostly young professional population, had tried and abandoned a usual GP system in favour of “everyone sees everyone”.

Many practices espoused continuity whilst describing systems that appeared to conflict with the goal of continuity—such as multidisciplinary clinics in which different elements of care are dealt with by different practitioners or appointments that are bookable only on the day. These arrangements had often been introduced to deal with rising demand, with loss of continuity as an unintended consequence.

Most practices operated a triage system for urgent appointments where access to care is prioritised over continuity. Some interviewees suggested that continuity was not as important as patients receiving the “right” care and felt that patients would inevitably move “from one to the other” practitioner, especially in urgent or high-priority situations. However, continuity was viewed as particularly important for certain patients (especially those with complex needs and multiple long-term conditions).

Westerly have undertaken work to improve continuity, and a GP from Fernleigh described continuity as having a “positive impact on workload” since “tasks are easier if the GPs know the patients well”.

Whilst many interviewees mentioned risks associated with digital technologies leading to threats to quality of care, we have to date identified few actual examples of these. Risks were most evident when clinicians talked about their attempts to do video consultations (which, as noted above, have been abandoned by most practices). In terms of quality, their concerns were threefold: challenges and time spent setting up the technology (including supporting the patient to do so), technology failure (perhaps due to human error), and poor image quality (reliant on the very variable set-ups which patients had at home—especially if they are using a mobile).

THEME 3: Meeting the needs of the disadvantaged and digitally excluded. A strong commitment to meeting the needs of disadvantaged groups was a universal core value across all participating practices. When asked to describe their practice’s ethos, one said it was to “engage with all sectors of the population” and another highlighted their practice’s emphasis on “engagement with hard-to-reach parts of the community”. Similar phrases were used by staff from ‘deep end’ practices in deprived areas and staff from practices in more affluent areas with postcode pockets of deprivation. In the former case, the entire practice logistics were oriented to serving a predominantly or exclusively deprived community; in the latter case, staff were keen not to orient towards the affluent majority at the expense of the more deprived minority (e.g. a traveller site or poor estate).

Those identified at particular risk of digital inequalities were people who were poor, elderly, homeless or in poor accommodation, those with drug or alcohol use problems or who speak limited English or lack full citizenship (e.g. asylum seekers), those who are hard-of-hearing (for phone consultations), or with learning difficulties, or with complex physical or mental health needs. The question practices wrestled with was how to align the aspiration to meet the needs of these groups with the reality of an increasingly digital service.

Disadvantaged patients often have complex needs, with multiple social and health problems (poverty, homelessness, low health and digital literacy, chronic illness, cognitive impairment) exacerbating one another. GPs emphasised that it was important to take a holistic perspective, investing in more challenging aspects of a deprived community: “Making a difference with the difficult stuff has a knock-on effect within the population.”

Practices were quick to identify various groups of people who find it hard to navigate the health system and emphasised the efforts to ensure that these patients are made welcome and able to get the care they need.

Access (addressed above) is a major component of the equity agenda—but access comes with an equity trade-off: an increasingly digital orientation makes the practice more accessible for some (often the young, digitally capable and less in need)—but at the cost of making it less accessible to others (particularly, the elderly, those not online and those with complex needs). Even when patients are (apparently) digitally connected, they may be unused to using digital technologies either at all or for their healthcare needs. One practice in a deprived locality (River Road) has been encouraging patients to use online consultations but found that some are not in the habit of checking their messages so may miss the GP’s reply.

As noted above, some practices effectively operated an ‘open door’ policy, allowing patients to walk in and book an appointment then and there. One practice (Ogden East), serving a deprived community, was described by staff as striving to be “very patient centred. Patients are taken seriously if they suggest their need is urgent even if the support staff feel that the problem is relatively minor.” In contrast, an interviewee from a practice with a more affluent population with some postcode pockets of deprivation (Newbrey) observed that the people who want to see a GP are rarely the ones who need to see one, hence being overly responsive to patient demand may paradoxically result in a less “patient-centred” service as the most needy are more likely to be overlooked.

Interviewees described how disadvantaged patients often lack family support and their social networks may be sparse—which means they may lack people they can call on to assist them with digital access. Staff commented that the increased social isolation that came with the pandemic had increased demand for in-person appointments (e.g. from young single mothers). On the other hand, assistance from a family member or friend brings its own challenges of privacy, confidentiality and failed demand.

Our initial interviews identified few examples of practices proactively helping patients acquire digital skills (they lacked the capacity to deliver such support), or of up-and-running digital navigator schemes. In some practices, receptionists sometimes helped patients complete their online consultation forms when they phoned in—though such activity may not be an efficient use of receptionist time. In practices with a high proportion of limited English speakers (Range Park, River Road), community navigators are already employed but it is not yet clear whether or how they are assisting in supporting digital access. One practice (Fernleigh) is considering piloting a ‘digital buddy’ system among its affluent elderly village population (which includes retired professionals), in which more digitally confident patients volunteer to link with and support less confident ones.

Some patients, especially those in the multiple jeopardy of several kinds of disadvantage, may be unable to access any services digitally. Staff in practices serving deprived communities noted that digital is increasingly the default option for secondary care and community services, “when the default method of access to a service is online, this is a potentially an illegal breach of NHS standards to ensure equitable access.” They sometimes needed to act as advocates for their patients to overcome digital barriers in other sectors (e.g. a mental health referral service in which the patient must complete a web registration form to enter the system).

THEME 4: Supporting and training staff and students. In the context of a wider staffing crisis affecting the UK NHS in general and general practice in particular (especially in deprived areas)²², a major contributor to staff morale is workload and the changing division of labour. As noted above, workload is high in all practices and some practices (Newbrey, Queens Road) feel it is at, or close to unsustainable levels. Interviewees attributed some of this workload to factors other than the move to digital (rising patient demand, unfilled staff posts, task-shifting from secondary care). But they considered some as resulting directly from digital options which, far from making work more streamlined and efficient, have made it less so. Online consultations in particular were widely viewed as stressful and inefficient.

Practices are taking various approaches to reducing the very high levels of workload. Attempts to closing the “digital floodgates” by reducing availability of appointments or restricting the time window of the online consultation service was covered under theme 1 above. Another approach is optimising the division of labour within the practice.

The larger practices in our sample (Camp St, Fernleigh, Newbrey, Towerhill, Westerly) have a high degree of disciplinary diversity including GP partners, salaried GPs, advanced clinical practitioners, physician assistants, paramedics, nurses, healthcare assistants, clinical pharmacists, dispensers, and a range of administrative and managerial support staff (including specific staff to support digital innovation in one practice). A wide range of additional staff often allows the GPs to be “freed up to take a more supervisory role” (Towerhill interviewee). This “freeing up” also releases the GPs to take on training roles, deal with the more complex and difficult cases, take on executive roles in running the practice and bringing in new innovations, or undertake more outward-facing responsibilities such as working on local medical political groups or practice consortia. Non-medical clinicians and GP trainees look after the more straightforward patients such as those needing long term condition management or with acute minor illness. Dedicated support staff – e.g., the practice manager or senior administrator—oversee the routine running of the practice. All GP partners, however, were still undertaking ‘normal’ day-to-day GP work as well, taking their turn in seeing emergencies and doing booked surgeries.

This kind of advanced division of labour in larger practices illustrates why an organisation’s size is such a strong predictor of its ability to innovate²¹. Apart from some community link worker roles, the smaller practices in Scotland and Wales (e.g. River Road, Range Park, Carleon), were less able to benefit from this multidisciplinary model of working—perhaps for historical or geographical reasons, or because there was less support to develop it in those regions. Carleon, however, has employed paramedics and a pharmacist from England working remotely.

Many interviewees described or alluded to low staff morale. In particular, they described all day remote telephone consulting as having a negative effect on their wellbeing. They missed the team element and seeing each other and the chance to interact with their patients in real life. There was a general sense that practices are currently in a state of flux—they feel that there is no going back to where they were before the pandemic but they have not yet reached a steady state (i.e. a way of running the practice that feels like a sustainable form of business-as-usual).

Some interviewees even depicted staff wellbeing as a safety issue. One GP, for example, started a discussion of future practice plans by saying that the most important thing was to “keep everybody safe – implementing the workflow plans to keep staff and patients safe” (practice name omitted to increase confidentiality). This interviewee described ongoing difficulties in staffing for GPs (salaried and partner), nurses and receptionist.

A national training lead expressed grave concern for wellbeing of trainees and young GPs who worked remotely from home and did not have the benefit of ‘corridor consultations’ to discuss concerning cases; one organisation had introduced ‘virtual coffee breaks’ where doctors working remotely could emulate this kind of mutual support and collaborative learning in the digital space.

Many of the problems of staff wellbeing in our participating practices centre around access and triage, in the sense that whoever is bearing the brunt of digital and wider access appears to be suffering (either receptionists are struggling with jammed overloaded telephone lines or GPs are wading through online consultations). Workload increased when patients over-use a system (e.g. sending in multiple requests because they have not yet heard back)—even for understandable reasons. One interviewee commented that remote consultations had “added a level of complexity” for receptionists, and in several practices (e.g. Westerly, Fernleigh), turnover of reception staff has increased.

Triage was one of the most contentious and problematic areas contributing to lowered staff morale and wellbeing. As noted above, systems for dealing with triage varied, and in several practices these were undergoing change. Some practices triaged patients as they telephoned the practice—working out when a patient needed face to face consultation or could have a phone call or an appointment with another service—whereas some used total telephone triage (i.e. call-backs) for all patients. Systems where a decision had to be made about modality sometimes seemed to generate staff stress and further problems down the line. Because of high demand, staff rarely had the time or headspace to channel the patient to the most efficient route to care. As one interviewee (in Fernleigh) said, “whereas pre-pandemic the team were trying to signpost to other available community services (such as the minor eye casualty service or minor injury unit etc) now they have mostly given up and just find any available appointment they can”.

At this stage, we have limited data on education and training in relation to increased use of digital technologies and services, although 8 of the 11 practices are training practices. A few interviewees described how students, trainees or early-career clinicians were finding it difficult to become skilled and confident with so much clinical practice happening remotely. With telephone consultations in particular, trainees felt they were very much “on their own”; the physical arrangements meant that they could no longer easily drop in on a next-door consulting room or start an informal corridor or tea-room conversation to discuss cases. Rather, specific plans had to be made for trainees to be able to debrief on patients. We will be exploring this sub-theme in more detail as the study unfolds.

THEME 5: Technologies and their associated infrastructure. Star defined infrastructure as “what other things run on” (including both technological components such as wires and servers, and also the human, organisational and regulatory ‘scaffolding’ that puts technologies in place and supports their use)²³. She observed that a feature of infrastructure is that it is generally backgrounded but becomes visible on breakdown. In this study, the wider technological and regulatory infrastructure was most noticeable by its relative absence—for both good and bad reasons.

In terms of policy, all practices in our sample appear to have been strongly encouraged and supported to adopt digital technologies and offer remote services during the pandemic. Development of digital access and digital consultations as a long-term strategy was largely in response to a national “remote by default” policy directive made in July 2020¹⁰, but a more recent reversal of this (pressure from the new Secretary of State for Health to revert to “in-person by default”¹¹) had a mixed reception from GPs and their staff, since it cut across the changes to digital services that practices had been working to achieve.

We detected widespread unease about what changes might be about to happen at the political and policy level (where, broadly speaking, GPs and their work is perceived as undervalued). In a letter to patients Queen Road explained that “A sustained attack by the media on general practice and the seemingly ill-informed demands of the health secretary for more face – to – face appointments have left us demoralised, broken and burnt out.” Some of the practices (especially those perhaps nearest to the policy process) expressed heartfelt concerns about the future of general practice more generally—with near-unsustainable workload, a workforce crisis and apparent government inaction. In Towerhill, one GP said that their involvement in the local primary care network “helps grow the standing of the practice,” suggesting that it acts as a “hedge” allowing the practice to deal with what politicians “throw at general practice.”

Health information infrastructures are also patchworked and path-dependent, in which components emerge incrementally and so cannot be installed or replaced wholescale²⁴. The varied fortunes of the 11 practices illustrate how pre-existing technological infrastructure (the ‘installed base’ in Star’s terminology²³) both enabled and constrained remote consultation and triage practices. Prevailing infrastructural arrangements led to the selection and use of particular technologies and the development of particular routines, processes, knowledge and workarounds, which in turn set the organisations on a particular infrastructural path going forward.

As noted above, most case sites initially introduced video communication platforms that had been specifically developed for medical consultations (Attend Anywhere and accuRx), but did not persist with this modality. Whilst these bespoke products have been designed to align with clinic workflows (e.g. ‘virtual waiting area’ to help manage the flow of patients attending their virtual appointments) and information governance requirements (e.g. avoid the need for patients to download software or provide personal information), implementation has been limited by network connection problems (at the practice and/or patients’ homes), difficulties interfacing with electronic records, a lack of adequate audio-video equipment and private space in the clinic, and the time involved to set up and troubleshoot the technology—all of which potentially jeopardise the professional standards of care, risk and workforce capacity described above.

Practices that have continued to use video are generally characterised by a strategic investment in IT and material infrastructure (e.g. clinic and office room set up, dual screens to view video alongside patient records), targeted use of the modality (with a clear understanding of how and when video would add value), local knowledge and skill to use and support each other with the technology, a degree of technical integration across video and electronic record applications (specifically accuRx and SystmOne), and the careful alignment of clinic workflows with software functionality (e.g. to book video appointment slots and support real-time video connection when deemed necessary by the clinician).

Whilst the telephone is an old technology, the extended use of telephone for triage and consultations depends on both traditional (‘legacy’) systems and also new or extended systems which (for example) allow patients to send digital photos and documents. Clinicians talked about how these technical adjuncts, alongside new clinical and communication skills, have reshaped their perceptions of the potential role of this medium in clinical care. Some sites have developed (and others are considering) advanced telephony systems, such as wifi connecting phones with headsets (Fernleigh) and phone call recording and cloud storage (Westerly).

Our initial interviews have highlighted the infrastructural work that has gone into creating and embedding new work processes and routines. Organisational routines are defined as “recognisable, repetitive patterns of interdependent action carried out by multiple actors”²⁵. Routines are situated within a socio-material context—in other words, the interdependent actions of human actors are structured around time, physical spaces, and material and technological artefacts²⁶. Participants described the challenges in distributing and coordinating administrative and clinical tasks, and the emergence of ‘hidden’ or ‘invisible’ articulation work (defined as work that is necessary for dealing with anticipated contingencies, but which is not formalised or documented²⁷), in order to support and accommodate the technology. For example, reception staff at Rhian routinely print out emails and other electronic messages from patients, and transfer the paper documents to a physical in-tray in the office, thereby aligning old and new systems of collaborative working.

Another example of articulation work is now receptionists in traditional practices such as River Road complete the online appointment request forms on behalf of patients (on the other end of the phone) who are unable or unwilling to use the online system themselves. Such work contributes significantly to ensuring that remote consulting and triaging practices are ongoing and feasible (‘keeping the show on the road’), though it is not known how much time is spent in this way nor how it increases or decreases the efficiency of the triage system in different settings. Our ongoing research will seek to understand this kind of articulation work in the busy setting of general practice reception areas and back offices.

In addition to the patient access and digital exclusion issues described in theme 3 above, staff interviews have also highlighted usability and access problems more generally. Sometimes staff have been able to address basic design flaws in subtle but important ways. For example, Westerly saw a significant increase in the use of online consultation requests after they updated the practice website to make it easier for patients to navigate and locate the electronic forms.

However, service teams are often unable to change or reconfigure the technical aspects of the system because they lack technical knowledge, IT support and relevant permissions. For example, many online consultation templates are considered too long and burdensome, with much redundancy and repetitive questioning, but these cannot be altered by the practice. A number of interviewees highlighted design flaws within their telephony systems for managing call queues. For example, in River Road, one of the main reasons for introducing the ‘Footfall’ online booking system was to address patient frustration (and clinical risk), as one nurse explained: “We don’t have any control over the phone lines. So people were phoning in, and there wasn't a message to let them know, like they were in the queue. It just rang and rang and rang. People weren’t aware that they were in a queue, and so they were just phoning and hanging up, because they did not think anyone was answering. They were getting very frustrated. They thought we were just sitting here, having tea and coffee and not answering the phone….”.

The availability and affordability of technology has also been shaped by funding and procurement decisions, including commercial contracts and professional standards. For example, Clinical Commissioning Groups (CCGs) in England have provided funding for GPs to use accuRx, and government-funded initiatives in Scotland and Wales focused on the roll out of Attend Anywhere in primary care, as part of the pandemic response. Procurement processes remain challenging, with a limited range of solutions that may not be fit for purpose at either end of the digital spectrum. Camp St Group are currently awaiting a CCG funding decision to use a new AI-driven triage and patient flow management system, called KLINIK. They are experiencing this process to be slow and uncertain.

THEME 6: Patient involvement in improvement efforts. Although some practices solicit feedback from their patient groups, our preliminary interviews have shown that in general these groups either do not exist or have a demographic (e.g. retired professionals) that is atypical of the practice’s population. There is only informal and ad hoc data on how patients are finding the new systems, the level of patients’ own technical capabilities and how they match the systems that are being offered. One or two practices are undertaking small-scale studies on this topic, and our own research includes a workstream on patient and public involvement.

Summary

Our in-depth case studies of a diverse sample of 11 general practices have illustrated both commonalities and differences in their approach to digital services. Practices vary in their enthusiasm for and uptake of such services. However almost all struggle with access and demand and with how to ensure that they are prioritising and meeting the needs of vulnerable and disadvantaged patients.

Themes to explore further

We will be taking forward some high-level issues (listed below) which have emerged from our previous work and our work so far in this study.

Patient input to practice change efforts. The limited input of patients to the design and evaluation of digitally supported services was striking, and due (we surmise) largely to pandemic-related restrictions on meetings and the high levels of workload and staff stress in many practices (there is simply no slack to undertake patient consultations). We hope that the co-design component of this study (described in our protocol paper²) will help to bring patient-centredness in improvement initiatives more to the fore.

Efforts to improve the triage process. Triage seems to lie at the heart of much concern over workloads, stress, staffing and staff morale. Several practices have recently changed their triage system and others plan to shortly. It is already apparent that there is no one-size-fits-all triage system, but we hope to tease out what is likely to work for whom, in what kind of circumstances.

Efforts to reduce inequalities. Practices are at an early stage in various efforts to support those who may potentially be excluded as services go digital. Our in-depth study design will enable us to explore intersectionality—how different social determinants (e.g. being elderly and poor and chronically sick) combine and interact to worsen digital inequalities. We will also be undertaking co-design activities using digital personas to support efforts to overcome these inequalities.

Quality of clinical care. Our early interviews did not pick up on much in the way of comments about what actually goes on in the consultation, nor many specific comments about quality and safety when managing long-term conditions, early indicators of serious conditions (which might be missed in the absence of an in-person encounter), and patients with communication challenges or complex needs). Previous research by ourselves and others suggests that remote care may compromise the therapeutic relationship and continuity of care, lead to more transactional forms of clinical interaction, fewer ‘doorknob consultations’, and delayed diagnosis of serious illness (see our protocol paper for literature review²). These problems are likely to affect the patient population disproportionately and generate new kinds of inequity.

Whilst remote assessment may have unacceptable risks for complex and vulnerable patients, it may be convenient and safe (and be associated with better uptake) for routine follow-up of patients with stable long-term conditions. However, there is a danger that if such reviews are undertaken by text messaging, the patient becomes (predominantly at least) an online entity, with adverse impacts on the therapeutic relationship and missed opportunities for key hands-on clinical checks (e.g. foot pulses in diabetes).

Selection and procurement of digital technologies. It was evident from our early interviews that some technologies introduced at the height of the pandemic have subsequently been abandoned because they were unfit for purpose and in some cases worsened the problems they were introduced to solve. The early pandemic was a time of relaxing red tape and bypassing regulatory approvals⁴, and governance (financial and clinical) now needs to be fully restored across the NHS. The procurement process for new technologies in the NHS is not always well-aligned with business cycles²⁸.

Technical functionality. It was clear from our interviews and other data-gathering that the ‘same’ technology (a telephone system, an online consultation system) can have very different functional characteristics depending on the precise product used, which functions have been enabled (or disabled), which local infrastructure it interfaces (or fails to interface) with, the demands placed on it, and human factors such as confidence, training and informal support to use it. With few exceptions, technological resources and know-how were greater in larger practices. As noted in theme 5 above, legacy infrastructure and contracts signed in the past sometimes created path dependencies which prevented practices from upgrading or replacing digital technologies in the ways they would have liked to. We will explore such issues in ethnography and digital walk-throughs as the study progresses.

Reverting to a more in-person model of care? As the pandemic recedes, practices are re-evaluating the benefits of the digital-by-default technologies and ways of working that they adopted in early 2020. Many are now in flux. They have clearly been through a huge change and a process of destabilisation, and are now searching for a sustainable way forward in the longer term. Again, there is no one-size-fits-all model but we hope to support and describe some ways of gaining an effective balance of traditional and digital forms of care.

Support for small practices. The apparent dependence of successful digital services on a sophisticated division of labour, and the latter’s dependence on practice size, raises important questions about critical mass going forward—either small practices are destined to become obsolete or different ways must be found to support them.

Planetary health. This theme did not come up in our early interviews for this study, but has featured in our previous research and we will be actively exploring it in future interviews. Travel to healthcare appointments generates greenhouse gases. Remote service provision could potentially reduce this, though carbon savings in primary care may be modest as patients live locally, and could be achieved at the expense of waste (e.g. over-diagnosis, over-treatment or over-referral). Local savings (of various kinds) may come at the expense of ‘hidden’ environmental waste.

Underlying data

Selected data on this ongoing, mainly qualitative study will be made available to researchers on reasonable request to the lead author. The reason we have not provided full transcripts for all interviews and copies of field notes is that the study design precludes this. We have carefully built relationships with each of the 11 practices based on personal contact from a researcher-in-residence, and worked extensively with staff to build trust and assure the confidentiality of information shared. Our raw data contains highly sensitive information (e.g. receptionists may be fearful that a GP or practice manager in their own practice might read negative things they have said; GPs may have voiced concerns about the commitment of trainees or vice versa). Whilst these raw data will inform our emerging understanding of each individual practice and also the cross-cutting analysis of all practices, we have an over-riding duty to the participants to keep these transcripts confidential. A breach of this duty would not only be unethical but could lead to the practice withdrawing from the study. Our NHS ethics approval is based on assurance of confidentiality of material disclosed by staff members and patients in the practices. For this reason, the only data available to be publicly shared is summaries of the practice familiarisation documents that have been approved by the practices. However, it may be appropriate for experienced researchers in this field to seek particular additional data from the corresponding author whose email address is given above, and any such request will be treated on its merits.

Extended data

Mendeley Data: Remove by Default 2. https://doi.org/10.17632/cx6v6zkp49.1²⁰.

This project contains the following extended data:

Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).