This chapter summarises the findings and presents a synthesis of the study. It draws on the discussions in the emerging-findings seminar that was held with participants in the study, patient and public representatives and advisory group members in December 2017. We first of all address the study questions:
- What is the extent of the adoption and deployment of PAs employed in hospital medical services?
- What factors support or inhibit the inclusion of PAs as part of hospital medical teams at the macro, meso and micro levels of the English health-care system?
- What is the impact of including PAs in hospital medical teams on the patients’ experiences and outcomes?
- What is the impact of including PAs in hospital medical teams on the organisation of services, working practices and training of other professionals, relationships between professionals and the service costs?
We then present a synthesis against a framework of dimensions to judge the quality of health services. We discuss the strengths and limitations of the study before presenting our conclusions and summary of recommendations.
The extent of the adoption and deployment of physician associates in hospital services
We found that PAs were increasingly being adopted into hospital workforces in England. A small but growing number of hospitals in England were employing PAs to work within a widening range of medical and surgical specialties (see Chapter 3, Survey of medical directors and Survey of physician associates). The growth was evident in comparison with data published for the UK in 20121 and reflects the data captured since in the 2016 UK FPA census.59 This pattern of small but steady growth in adoption reflects those reported in the USA, the Netherlands and Canada.22,143,144 We report for the first time that when some PAs have vacated their posts in secondary care they have not been replaced owing to the shortage of PAs to employ. This situation is likely to change with the public-finance-supported growth in training (through HEE) resulting in an estimated 3200 PAs in the UK (mostly England) by 2019.145
We found that PAs, with the exception of those employed in emergency medicine, were mainly deployed to undertake inpatient ward-based activities of the medical/surgical team during weekdays, mostly in the daytime between 07:00 and 19:00 (see Chapter 3, Survey of physician associates, and Chapter 5, The physician associates’ working patterns: times and places). We found only a small number of PAs spending any time in outpatient clinics and theatres, and when they did it was a minor proportion of their working hours. This finding has also been reported in the USA146 and the Netherlands.147,148
It was evident that while there was a core role in inpatient care, many of the individual PA roles were ‘bespoke’ or ‘moulded’ in that they were developed/trained to meet the requirements of an individual medical/surgical team, but their activities also reflected the experience, capabilities and competencies of the individual PAs (see Chapter 5 and Chapter 7). The longer a PA stayed in a post, the more skilled and knowledgeable they were reported to become in the work of that specific medical team and specialty. This was contrasted with the greater breadth of knowledge and skills developed by early-career doctors through rotation to different specialties. We found that some PAs had been trained by their consultant(s) to undertake specific clinical procedures in support of faster patient access to these procedures. All grades of doctors described a process of building trust in individual PA’s competencies; this reflects the concept of ‘entrustable professional activities’, which is widely used in medical education literature.149 Some PAs had additional roles, such as a leadership role for other PAs, mentoring student PAs and hospital-wide involvement in specific initiatives related to quality improvement and patient safety. In addition, some had part-time roles within universities as PA educators.
The factors that support or inhibit the inclusion of physician associates as part of hospital medical teams at the macro, meso and level of the English health-care system
At the macro and meso level of the English health-care system, we found government policy and regional health plans supporting more advanced clinical practice roles, including PAs, in order to address medical workforce shortages (particularly in general practice) and the growing demand for health-care services (see Chapter 3, The policy review). This identifiable policy support was in contrast to the situation at the beginning of 2014, when we noted that PAs were absent from all relevant English policy documents, although they did feature in Scottish policy documents, particularly in relation to rural and remote areas.17 The policy support identified in this study was more than rhetoric, with public finance assigned to supporting training with a target number of PA graduates, albeit with particular reference to general practice rather than acute hospital service (see Chapter 3, The policy review). By 2017, there were 33 PA studies courses in the UK,150 compared with two in 2013.17
Our national survey of MDs in England indicated that a growing number were open to more advanced clinical practice roles, including PAs, to address junior doctor workforce shortages and increased patient demand (see Chapter 3, Survey of medical directors). However, the small pool of PAs to recruit from was identified as an inhibiting factor and PAs were not necessarily perceived as the right workforce solution by all MDs or professional leaders. Another important inhibiting factor that was reported was the lack of regulation of the PA role and concomitant lack of authority to prescribe and order ionising radiation. A public consultation was led on this issue by the Department of Health in late 2017 providing further evidence of increased macro-level support in the system. The results of the consultation were finally published in February 2019 with the decision to include physician associates in state regulation.151
We found that the macro- and meso-level factors enabling and inhibiting the inclusion of PAs in the workforce were amplified and expanded at the micro level of the hospitals (see Chapters 4–9).
At the micro level, it was reported that developing advanced clinical practice roles, such as PAs, was a necessity in the face of the shortages of junior doctors to cover the medical rotas; the need to release junior doctors to undertake their training; the increased workload created by increased patient demand and expansion of services and recognised quality issues in service delivery.
Many different types of stakeholders supported the recruitment of PAs as a new source of advanced clinical practitioners in preference to depleting other staff groups, in particular nursing, which was also experiencing severe shortages. In addition, the PAs’ training in a medical model was reported to be a good fit within the medical/surgical team for undertaking the required medical work. Evidence from North America suggests that similar factors have supported the inclusion of PAs in acute medical/surgical teams.133,152,153
As part of an innovation and change process,154 many participants within the hospitals identified the benefit of having clinical leadership, ongoing clinical champions for PAs and visible executive-level support (see Chapters 4, 7 and 9). This has previously been described in relation to one acute hospital unit in England.44 The successful contribution of PAs in one or more consultant teams was identified as leading to other teams creating posts for PAs (i.e. the spread of innovation through diffusion51 rather than a top down managerial-led change). We return to the contribution of PAs in the following section.
Against these supporting factors, inhibiting factors were also identified at the micro level. Foremost among these was the lack of role regulation with attendant lack of authority to prescribe or order ionising radiation. The lack of regulation raised concerns about governance, responsibilities and liabilities (see Chapters 4 and 7), issues that have been raised in other countries where PAs have been developed or piloted and as yet do not have regulation or licensing procedures.155,156 The extent to which the lack of authority to prescribe or order ionising radiation was viewed as an inhibitory factor varied between specialties.
A few clinical leaders (medical and nursing) were reported to have views that PAs were not the right group to develop or employ, favouring nurses or others. Likewise, some consultants working in some specialties with high-dependency patients, having employed PAs, were not re-appointing when PAs left as they considered doctors to be a better fit to the particular needs of that team and patient group. The extent to which the lack of prescribing or ordering ionising radiation authority, rather than other expressed concerns, was a factor in this experience was not clear.
Overall, the view was that PAs would become much more useful and attractive to the medical/surgical team if they had prescribing authority and there was much puzzlement as to why they did not have these rights. The Netherlands, where the first PA programmes commenced in 2000, agreed the legislation to regulate PAs and give them prescribing authority in late 2017.157
Finally, the lack of a pool of PAs to recruit from was considered a serious inhibition and this was linked to concerns as to how best to retain PAs in medical/surgical teams. Other evidence from surveying PAs in the UK has suggested that retention might be an issue in some places as some PAs did not consider that they were being effectively used within the clinical team.158
The impact of including physician associates in hospital medical teams on the patients’ experiences and outcomes
From interviews with inpatients and family members in this study, we found that they perceived the PAs as important in keeping them informed about their medical care and management plan (see Chapter 5). PAs were reported to be caring, approachable and good at communicating. Patients were very pleased with the explanations of important information they received about key aspects of their care. There was an implication that they better understood the explanation from the PA than the more complex language used by some doctors. PAs were trusted within their teams to convey important information to patients. Patients were content, and felt that it was appropriate, to be physically examined and have basic medical procedures carried out by a PA. Overall, inpatients and relatives reported being very happy with the care they received from the PA and the hospital staff.
Patients and relatives did not understand the PA role as a concept, but placed more emphasis on receiving good care, rather than the job role or title of individuals in the team treating them. This contrasted with the patient and public voice representative groups, which considered that it was important that patients understood exactly who and what members of their clinical team were. All participants were happy to have a PA involved in their care in the future. Some additional caveats were added, such as they thought that PAs were a good idea (to assist with staffing pressures) as long as they were properly supervised. In addition, some participants wanted reassurances that this addition to the medical team did not result in fewer doctors. High levels of satisfaction with PA involvement in acute inpatient care for similar reasons has been reported from studies in North America and the Netherlands.109,110
All types of staff (including in emergency medicine) interviewed across the case study sites reported that patients really appreciated the work of the PAs, as evidenced through positive comments, compliments, gifts and seeking them out to talk to (see Chapters 4 and 6–9). We discuss the patient outcomes in more detail in the following section.
The impact of including physician associates in hospital medical teams on the organisation of services, working practices and training of other professionals, relationships between professionals and the service costs
We consider first the impact on organisation, working practices and costs, before turning to the impact on relationships and on the training of other professionals.
The impact on organisation, working practices and costs
The systematic review found 16 observational studies in the specialties most frequently employing PAs in the UK (see Chapter 3).109 They were all single-site studies and of variable methodological quality. In four studies of emergency medicine and one study of trauma and orthopaedics in which PAs were an additional resource to the medical team, there were reported reduced waiting and process times, lower charges, reduced re-attendance and admission rates and good acceptability to staff and patients. Analgesia prescribing, operative complications and mortality outcomes were variable. In internal medicine, outcomes of care provided by PAs and doctors were equivalent.
In synthesising the evidence from our study, we draw from our findings reported across Chapters 4–10. We found that, primarily, PAs were deployed to help address gaps in medical rotas. Gaps in medical rotas in hospital teams have been reported as an issue of concern in the NHS.159,160 There was variability in whether PAs were in addition to or part of the medical rotas. In some services, PAs were part of staffing rotas for what was described as the second-tier staff. This was referring to doctors below registrar level and/or advanced clinical practitioners. In some services, the PAs’ duty times (rotas) were arranged to cover for absences of doctors (e.g. to attend training) and reduce the use of locum doctors. Consultants and operational managers considered locum doctors who were new to their service as less efficient, less safe and more costly than PAs. Reduction in the use of locum doctors, particularly in reaction to costly agency fees, has been a major issue for the NHS, with spending caps imposed in 2016, just before we began the data collection for the case studies, and still in force.161 However, it was evident that decisions about employing and deploying PAs were not just about finance but also about enhancing the quality of service provision and improving support to junior doctors in training. PAs were substituting for some of the service contribution of the early-career grades of junior doctors but there was variation in the extent to which individual PA posts were substituting for and/or supplementing30,52 the work of junior doctors.
One of the most frequently reported impacts on organisation was that PAs provided continuity of staffing in the medical/surgical team. This, in turn, was seen to provide benefit to other staff and patients through:
- Continuity in presence on the inpatient wards, thus increased access to the medical/surgical team for patients and nurses.
- Continuity in knowledge about current inpatient status, management plans and patients’ progress, thus facilitating updating patients and the medical/surgical team, for example on ward rounds.
- Continuity in knowledge about the policies and practices (clinical and otherwise) of the department, the individual consultants and the trust; this was of particular value for new doctors to that particular workplace.
Similar observations about PAs providing continuity within the medical/surgical team have been made in North America and the Netherlands.144,146,147
The PAs were reported to work alongside junior doctors in managing the medical/surgical teams’ workloads (although in some specialties the PAs’ full potential was not realised owing to a lack of authority to prescribe and order ionising radiation) and, for doctors in training, PAs were invaluable in induction into policies and working practices. Doctors in training in the UK have described the stress they face in understanding information technology systems, policies and procedures when starting new posts.162 PAs undertook significant amounts of non-patient-facing clinical work for the medical/surgical team (e.g. preparing discharge summaries). Similar findings have been reported in North America and the Netherlands.144,146,147 A systematic review of studies of physician time use in hospitals settings found that activities that were indirectly related to a patient’s care used more of hospital physicians’ time than direct interaction with hospitalised patients did.163 The extent of the patient-facing clinical work delegated to the PAs varied, dependent on the views and attitudes of the senior doctors in the team, as discussed in section The extent of the adoption and deployment of physician associates in hospital services.
The presence of a PA in the team was considered to release the doctors’ time in two ways: first, so that the doctors could attend more complex patients (irrespective of specialty); second, where PAs worked in inpatient settings, it released doctors to attend or remain with patients in outpatient departments and theatre (which was also required training experience). The presence of PAs enabling doctors to undertake required training was also reported and is consistent with reports from the USA.35
All consultants, registrars and managers reported the PAs to be safe, with no serious incidents or patient complaints recounted. This reflects the evidence found in the systematic review,109 as well as in a Dutch study of PAs in other surgical specialties110 and in a recent single-unit survey in England.164
Doctors of all grades described a process of establishing an understanding of the PAs’ competency and of building trust in their capabilities through working with them. We have previously referred to the process of entrustable professional activities within medicine149 and we described this process in relation to GPs and PAs in primary care in England.18
We found that PAs were reported by all stakeholders to contribute to efficiency by smoothing and improving patient flow into, during and out of an episode of treatment/care. PAs were considered to provide ‘oil’ to the system. The extent to which PAs supported the smoothness of patient flow was linked to other attributes of the PA role, discussed previously, such as bringing continuity to the medical team, undertaking non-patient-facing clinical work and providing a communication ‘bridge’ between the medical team, the nursing team and those in roles responsible for patient flow, often nurses. We return to this reported ‘bridging’ role issue in discussing relationships in section The impact on the relationships between professionals and the training of other professionals. We report on the PA support to smoothing patient flow for the first time in the English hospital setting and consider that it needs further investigation in the context of the variety of infrastructure and other support to the medical/surgical team found in NHS. The variability in infrastructure and staffing between NHS hospitals has been well established.165
All senior managers and clinicians described the difficulty of attributing patient outcomes and cost to an individual professional when clinical provision was team based and affected by multiple other contextual factors. Some thought that the presence of the PAs enabled the senior doctors to be more efficient. Many managers and consultants reported that to some extent PAs reduced the use of expensive locum doctors. Understanding current rates of pay for locum and agency staff in England is not straightforward. Although NHS England imposed caps on agency fees and hourly pay rates in 2016,161 there is considerable evidence that the shortage of doctors has resulted in high demand, which has driven these rates up.166 From a survey of 68 NHS trusts, the average hourly rate of pay to locum FY2 doctors in 2017 in England was reported to be £50.90 (the maximum reported for the period was an hourly rate of £104.69) plus an average hourly commission rate to the supplying agency of £5.60 (the maximum rate reported was £13.49).167 Validated national unit costs for staff, such as PAs, on Agenda for Change terms and conditions give hourly rates (inclusive of oncosts and overheads, but excluding qualifications) of £45 (band 6) and £54 (band 7).89 However, although reducing financial spend on locums was reported as important to the consultants and managers, the primary consideration given for preferring PAs to locum doctors was patient safety and efficiency.
None of the managers or clinicians, in any of the sites, was able to provide any routine data or reports from which the impact of the involvement of PAs could be considered. This absence of data has also been reported from surveyed executives of medical organisations in the USA, who noted the contrast to their primary care services, and similarly could not provide information on patient outcomes or costs that could be attributed to the inclusion or exclusion of PAs in hospital services.133 Our systematic review did not find robust evidence on cost-effectiveness of the inclusion of PAs in the specialties in which PAs are more frequently found in the UK.109 A recently reported large, matched, controlled, multicentre study from the Netherlands of several different surgical specialties found no difference in length of stay or quality-adjusted life-years for inpatients of doctor-only teams compared with those treated by mixed doctor and PA teams; the involvement of PAs reduced personnel costs but not overall healthcare costs.111
The pragmatic retrospective comparison of consultation records for patients seen by FY2 doctors or PAs in the ED found no difference in the primary outcome of rate of re-attendance. A subsample of records of patients seen by FY2 doctors and PAs were reviewed by senior clinicians and judged as equally safe and appropriate. We provide this non-inferiority evidence for the first time in the UK setting.
The impact on the relationships between professionals and the training of other professionals
A number of participants described initial resistance to the introduction of PAs from doctors and nurses. In most services, this was reported to dissipate over time as a result of actually working with PAs (see Chapters 4 and 7–9). Although positive relations were described between PAs, doctors and nurses, there were also occasional reports of doctors and nurses who remained less positive to the introduction of a new professional group. Abbott53 has described the way in which health-care professions are part of an interdependent system in which they jostle over areas of jurisdiction linked to issues of status and control. The dissipation of tension between professional groups through the process of working together has been described both by Abbott53 and in other studies of the introduction of new roles and advanced clinical practitioners,168,169 as well as in our study of PAs in primary care in England.18
Many stakeholders described the PAs as providing a ‘bridge’ between medicine and nursing in some of the inpatient settings. This was also a finding from our study in general practice.18 In the inpatient setting, the ‘PA as bridge’ was described as a mechanism for ensuring speedier access for nurses to the medical/surgical team, particularly in relation to patients whose condition was changing or in assisting in planning for their management or discharge (i.e. supporting patient flow). PAs were also described as accessible to others in the nursing team, such as health-care assistants, as well as for non-clinical staff.
Some early-career doctors were reported to be concerned that the presence of PAs would reduce their opportunities for training in certain procedures, as consultants would favour the PAs to do these. Most doctors and PAs described the prioritising of training for doctors and were alert to such problems. To the best of our knowledge, we report this for the first time in the UK setting.
Synthesis
This study is framed by the dimensions proposed by Donabedian49 and Maxwell50 for evaluating health-care services and is applied here to the innovation of PAs in secondary care. The dimensions are acceptability, effectiveness (including safety), equity (fairness) and efficiency (including costs).
Physician associates were an acceptable group of health professionals to contribute to secondary care at the macro, meso and micro levels of the health system. The macro and meso levels of support are in marked contrast to the situation when the primary care study was completed. PA training and employment is now supported at the government level by all countries in the UK.
Physician associates were found to be acceptable and appropriate by most, but not all, doctors, managers and nurses. Patients had little knowledge of the PA role itself but viewed the PAs very positively within the context of the medical/surgical team. Some consultants in specialties with high-dependency patients reported that doctors were more appropriate for the level and/or pace of work, particularly prescribing, required in that setting. Clinicians and managers were supporting the development of many types of mid-level advanced clinical practice roles, of which PAs were one. For many doctors, the training of the PAs in the medical model made the PAs particularly appropriate for the work they required. It should be noted that other hospitals in England have developed medical assistant roles (sometimes called doctors’ assistants or physician assistants) to support doctors. These are trained to the level of health-care assistants to undertake tasks such as phlebotomy and clerical work under the direction of the medical/surgical team.170 The contribution of different types of roles supporting the medical/surgical team requires further investigation.
For junior doctors and some nurses, there was uncertainty as to the scope of the role, an uncertainty compounded in the eyes of many by their lack of inclusion in the regulatory process for health professionals. Inclusion of PAs in these processes for health professions was viewed as important in terms of their acceptability, their perceived effectiveness (particularly in relation to prescribing and ordering X-ray investigations) and governance.
Physician associates were deployed, within a medical or surgical team, to mainly work in the inpatient ward under the supervision of consultants and their registrars. In the EDs, they worked as part of the team, with early-career doctors and other advanced clinical practitioners, and under senior medical supervision. Their work was described as aiding patient flow and releasing doctor time for more complex, medically unstable patients, for outpatient consultations, for the operating theatre and for training activities. There was no indication that this was inequitable (unfair) to any group of patients.
As part of the medical/surgical team, PAs were reported to be clinically safe in the work they did and to add to patient safety through their knowledge of policy and the practices of the consultants and the organisation. These colleague reports were supported by the finding that the patient outcomes and consultation records of FY2 doctors and PAs in the ED were equally safe and appropriate.
There was difficulty of attributing patient outcomes and cost to an individual professional when the provision was team based and affected by multiple other factors, but, in general, the reported contribution of PAs was highly positive.
Planning and developing a medical and surgical workforce is recognised as internationally challenging.145,169,171 Medical staffing shortages have continued to rise in England, with a government response during the study period to create an additional 1500 medical student places in England from 2018.172 The General Medical Council, although welcoming this growth, noted that their presence will not be felt in the workforce as graduates until 2023 and raised questions as to the availability of sufficient capacity within the workplace to train increased numbers.173 Cyclical shortages of different grades of doctors with attendant reports of increased pressures on the employed doctors has been a recurrent feature of the NHS.174 Achieving a sustainable medical workforce with the right balance of consultants, GPs, specialty doctors (i.e. not in training posts) and junior doctors has to take account of (1) the service demands, (2) the training requirements for career advancement and (3) the creation of manageable jobs within interesting careers. Having a cadre of flexible, mid-level practitioners, trained in the medical model, who can support doctors, but are not doctors, or intending to replace doctors, may help address inherent tensions between service demands, training requirements and budgetary constraints, as well as mitigate cyclical shortages.
Strengths and limitations
This was a mixed-methodology study that had multiple elements. The major strengths are as follows:
- the use of multiple methods over a number of workstreams, allowing a macro-, meso- and micro-level view of the research questions
- undertaking micro-level investigation that included hospitals from a wide range of different sites (both major conurbations and smaller cities/towns, large tertiary referral and small ‘District General Hospital-type’ sites), from a wide range of specialties and with a considerable proportion of the known PA workforce in secondary care in England
- collection of extensive qualitative data from a wide range of stakeholders utilising individual interviews, allowing analysis of perspectives gathered from senior managers (MDs and trust board members); members of the medical team (senior and junior doctors) and other clinicians, such as nurses; patients; and relatives
- gathering of quantitative and qualitative data on PAs ‘in action’ in hospitals, using both observation and self-completed diaries
- carrying out a well-powered quantitative comparative analysis of the documented processes and outcomes of patient care by PAs and FY2 doctors in three EDs in different parts of the country.
The main limitations of the case study elements were largely a result of the difficult circumstances and constraints on research within a highly complex and dynamic system, which was simultaneously under great pressures.175–179 The protocol was changed twice to try to accommodate real-world changes and priorities. For example, we had intended to compare activity of PAs and FY2 doctors while managing minor cases in EDs but found changes (between the time period we negotiated the research funding and when we commenced recruitment) in how many PAs worked in EDs and in what area of the ED PAs worked in, with most currently working in ‘majors’. This required a change to comparison of ‘majors’ cases instead. This change to ‘majors’ patients also introduced the potential for selection bias in the sample, in that patients in the minors section are attended by the next-available professional whereas patients in the majors section are attended by professionals based on a range of factors, such as severity/immediacy of the injury/illness, number of patients arriving at that time point and the availability of staff, as overseen by a senior doctor in charge. Although there might be the potential for selection bias by a senior doctor in charge in the majors section, three factors in the design mitigated this: (1) the use of three different ED departments rather than a single ED, (2) the overall sample was obtained from a 16-week period at all times of day and night in which there will have been the greatest variation in both the factors referred to above and also individual senior doctors in charge and (3) statistical adjustment for MTS,87 alongside other potential confounding factors, in modelling. The changes also resulted in a very limited economic analysis, unlike the analysis that was planned. There are still relatively few PAs working in secondary care and there is a smaller number within that group who are willing to volunteer to participate in this type of research. However, we successfully recruited six hospitals with PAs (and three with PAs in the EDs) and, within those, 41 PAs willing to participate.
We were unable to access quantitatively measurable data on the effect of the introduction of PA on patient outcomes (other than in the ED) or costs as the hospitals were unable to supply these type of data. In addition, the attribution of changes in quality of patient care, patient safety events or cost savings to the introduction of PAs is problematic when such an introduction commonly went alongside other changes or responses to other significant events, such as unprecedented surges in demand and workforce shortages. However, we were able to describe and scope the range of potential impacts, which could inform the design of future studies either through matched comparisons of medical/surgical teams with and without PAs, such as those used in the Dutch study,110 or in a ‘stepped-wedge’ design (in which the change is introduced sequentially in all sites so that all participants get the intervention, but not simultaneously).180
Conclusions
Physician associates are a new group of health-care professionals being adopted in a minority but growing number of secondary care hospitals across the UK. The well-documented workforce shortages in the NHS meant that PAs were one of many solutions being followed to address, in particular, the shortage of doctors. PAs were deployed to undertake some of the work of junior doctors, releasing their time to focus on complex and new patients as well as training requirements. The work that PAs undertook was mainly the ward-based activities of the medical/surgical teams and included significant amounts of medical administration. This work was considered to significantly aid patient flow and patient experience. The stability of PAs in the work setting, in contrast to doctors in training, was reported to be valued in inducting new doctors and also contributing to patient and junior doctor experience, patient safety and throughput and the development of trust between PAs and the medical and multidisciplinary teams. Patients were very satisfied with PA care as part of their medical/surgical team but had little understanding of the role. The pragmatic comparative review of ED records of patients attended by PAs and Foundation Year doctors found that they provided equivalent, safe care, suggesting that PAs are another source of advanced clinical practitioners that can be employed in that setting. The main cost benefit identified, but not quantified, by managers and clinicians was thought to be the reduction in employing locum junior doctors. Even when this was discussed, patient safety and experience were the main considerations. Although some specialties with high-dependency patients reported that PAs were less appropriate than junior doctors, most specialties reported inefficiencies and problems in the workflow created by the lack of PA authority to prescribe and order ionising radiation. It is clear from our study that PAs could provide a flexible addition to the secondary care workforce without drawing from existing professions or reducing opportunities for doctors in training. However, their utility in this setting is unlikely to be fully realised without the appropriate level of regulation with attendant authority to prescribe medicines and order ionising radiation within their scope of practice.
Implications
Physician associates have the potential to be a competent and flexible addition to current and future medical and surgical teams. Patients, the public and secondary care professionals need more information to understand both the role and also its potential contribution to acute care medical and surgical teams. Individual employing NHS organisations need to consider their role in this alongside that of national NHS and professional bodies.
Physician associates are a workforce innovation that is not necessarily welcomed by all other professional groups. As an innovation, successful introduction and growth is likely to depend on a planned process with attendant senior levels of support, local champions and embedded governance and review processes. This requires employer organisations to both plan and identify appropriate people and resources. With the expected exponential increase in PAs in England in the next 2 years, this requires prompt consideration.
Findings from this study suggest that the efficiency and contribution that PAs could make to medical and surgical teams in secondary care will be increased by the legal authority to prescribe, which can be achieved only by including PAs in the regulatory framework for health-care professionals. This has implications for policy with regard to regulatory frameworks for health professionals in the UK.
In order for there to be a supply of PAs for secondary care, there needs to be cognisance of this group in health profession education commissioning at a local level. Education commissioners at local levels should consider the contribution PAs could make to a future flexible secondary care, and not just primary care, workforce.
Further investigation
This study has highlighted a number of questions that require further investigation in the context of increasing numbers of PAs and other mid-level practitioners available in the UK to join the NHS workforce. These include:
- What are the patient outcomes, experiences and service costs for same-specialty medical/surgical teams before and after the introduction of PAs using stepped-wedge quantitative research designs?
- How does the work of the medical and surgical teams in the same specialties vary between hospitals with different types of infrastructure, technology and patient flow systems and what workforce is required to meet the variation?
- What is the extent of the use and role of medical assistant roles in different medical and surgical specialties and how does that compare and contrast with the use of advanced clinical practitioners, including but not limited to PAs?
- What is the influence of sociodemographic and geographic factors, and types of hospitals, such district general or tertiary hospitals, on the demand for PAs and other advanced clinical practitioners, in a secondary care workforce?
- What factors influence clinician and manager decision-making as to staffing and skill mix in secondary care teams, with particular reference to beliefs, preference, evidence and use of financial data?
- Can workforce modelling of different permutations of skill-mix support to medical and surgical teams, in which supplementation and/or substitution are explicit, assist managers and clinicians in workforce decision-making?
- What opportunities are available to retain PAs (and other advanced clinical practitioners) longer term in an essentially flat career structure?
- What impact will the rapid growth in PA numbers in England over the next 3 years have on perceptions and development of the role?