Patients’ experiences of consultations with physician associates in primary care in England: A qualitative study

Abstract Background Physician associates are new to English general practice and set to expand in numbers. Objective To investigate the patients’ perspective on consulting with physician associates in general practice. Design A qualitative study, using semi‐structured interviews, with thematic analysis. Setting and participants Thirty volunteer patients of 430 who had consulted physician associates for a same‐day appointment and had returned a satisfaction survey, in six general practices employing physician associates in England. Findings Some participants only consulted once with a physician associate and others more frequently. The conditions consulted for ranged from minor illnesses to those requiring immediate hospital admission. Understanding the role of the physician associate varied from ‘certain and correct’ to ‘uncertain’, to ‘certain and incorrect’, where the patient believed the physician associate to be a doctor. Most, but not all, reported positive experiences and outcomes of their consultation, with some choosing to consult the physician. Those with negative experiences described problems when the limits of the role were reached, requiring additional GP consultations or prescription delay. Trust and confidence in the physician associate was derived from trust in the NHS, the general practice and the individual physician associate. Willingness to consult a physician associate was contingent on the patient's assessment of the severity or complexity of the problem and the desire for provider continuity. Conclusion Patients saw physician associates as an appropriate general practitioner substitute. Patients’ experience could inform delivery redesign.

is the physician associate (PA), previously known as physician assistant, in England and the wider United Kingdom (UK). 4 The physician assistant role developed in the United States of America (USA) in the 1960s with over 86,000 PAs employed in all health-care settings, including primary care, in 2015. 5 PAs are trained in the medical model to diagnose, treat and refer autonomously, as agreed with their supervising physician, in line with local legislation. 5 Building on the model from the USA, PAs have been introduced to other health-care systems such as Canada, Australia, the Netherlands, Germany and India. 6 In the UK, the first PAs employed in the mid-2000s in the National Health Service (NHS) were American-trained. 7,8 The first UK-trained PAs graduated from post-graduate diploma courses in 2009. 4 Unlike PAs in the USA and the Netherlands, those in the UK do not currently have the legal authority to prescribe and do not currently come within a state regulatory framework for health professionals. 9 Concern about current and predicted shortages in the general practitioner (known in some countries as family physician) workforce, together with a policy emphasis of greater delivery of care outside of hospital, has led to recommendations for more PAs to be employed in primary care 10 and a policy statement by the Minister of Health in England that 1,000 PAs will be employed in general practice by 2020. 11 PAs are a recent innovation in UK general practice settings, and they have been mainly deployed to provide consultations to patients requesting urgent or same day appointments. 12,13 PAs in this setting are formally defined as dependent practitioners to the general practitioner, but can work independently in the practice health-care team, seeing and referring patients on and reviewing clinical test results. 4 A review of evidence regarding PAs in primary care from 1950 to 2010 found only six published studies from the United States which sought the views of patients who had consulted PAs. 14 Of these, five studies used surveys and reported high levels of satisfaction. [15][16][17][18][19] Within the UK, two short-term pilot schemes to introduce US-trained PAs to different types of services, including primary care in the NHS in England and Scotland, also reported high levels of patient satisfaction. 7,8 An observational study in England comparing PA and GP consultation records (n=932 and n=1,154, respectively), with a linked patient satisfaction survey (n=490 and n=590, respectively), conducted by the authors, also found that the majority of respondents were satisfied or very satisfied with their consultation with both PAs and GPs, and all but a very small number reported confidence and trust in the PA or GP. Eleven patients (4.1%) reported they would prefer to see a GP in future. 20 The conceptual issues and limitations of patient satisfaction surveys are well documented. [21][22][23] Satisfaction is a relative concept, based on evaluative judgements, 23 and in the instance of such a role innovation as PAs substituting for GPs, it requires more in-depth understanding of the dimensions upon which the judgements are being made. 23 Calnan suggested that a conceptual framework for lay evaluation of health care should include elements of the level of experience of health care and the goals of those seeking such care. 24 In-depth information about patient experience can be captured using interviews. 25 (p9). However, only one study which sought the views of patients who had consulted PAs, conducted in the USA, used interview techniques. 26 This study reported mixed responses from patients in an area where the PA had been the sole primary care provider for the previous two years, with the patients suggesting that they would sometimes prefer to see a doctor due to a) not having confidence in the PA (not being a doctor), b) already having a doctor or c) having a long-term condition requiring specialist care. 26 Against this background, our study addresses the evidence gap regarding the patients' perspective on the innovation of PAs providing general practice services, in a country where nurses are an established part of the state funded, general practice team. 27 The study draws on the interpretative tradition 28 and builds on our patient survey respondents' evaluative judgements to address questions of how patients understood the role of PAs and their experience of health care provided by a PA as a mid-level health practitioner.

| METHOD
The data reported here are from a larger study which involved six general practices employing PAs across southern England and six matched practices which did not. 29 The practices were purposively sampled to represent the different types of practice found in the UK by list size and number of practice partners, in urban and rural settings with varying levels of deprivation. 29 Five of the practices employed only one PA, the sixth employed two; four PAs were female and three male; four had trained in the USA and three in England.
Adult patients (n=430) were given a patient satisfaction survey, which included a request to volunteer for an interview, by reception staff as they left a same day or urgent consultation with a PA.
Completed volunteer forms, with contact details, were returned to the researchers. A topic guide was developed to explore issues not captured by the patient survey, that is patient choice about whether they saw a PA or not and their level of satisfaction with that experience and associated reasons; the patient's understanding about the PA role, exploring information provision and experience of seeing PAs; their experience of the PA consultation compared with their expectations of consulting a GP, probing issues of confidence and trust; how issues such as making a referral and prescribing were handled by the PA and the impact of this on the patient's experience; and their perspectives on consulting a PA and/or GP in the future.
One hundred and fifty-two patients expressed an interest in volunteering for an interview as part of the qualitative study we report here. Of these, contact details for 43 were incomplete, 40 did not respond to the researchers' contact attempts, and four contact details were received after recruitment had closed. Researchers made contact with 40 patients and, of these, 34 participated in an interview (all but one by telephone). Interviews lasted between 10 and 20 minutes. Four interviews were not used when it became apparent that the consultation being discussed had not been with a PA or the adult participant described a consultation for a child. With consent, the interviews were digitally recorded and transcribed. Interpretive analysis was conducted using thematic analysis 30 by two authors (LJ and MH) with another researcher. Transcripts were read and re-read; initial codes were developed through discussion and applied initially to a small number of transcripts, enabling further discussion and iteration of the thematic index. Coding against the index was undertaken by the same three researchers, with at least two carrying out parallel coding of each transcript. Any disagreement was addressed through discussion and further iteration to the analytical process if necessary. QSR International's NVivo 10 Software was utilized in the analytic process.
The study was approved by a UK NHS Research Ethics Committee.

| Description of participants
The thirty participants were unevenly spread across the practices (minimum two, maximum 11 per practice) but were diverse in terms of gender (12 female and 18 male), age (range from 27 to 90 years), ethnicity (nine people were of black and minority origin and the remainder were white) and socio-economic background as defined by the Index of Multiple Deprivation for their general practice 31 (see Table 1).
One participant was a carer who had accompanied their relative to the PA consultation. The types of health condition described by the participants in consulting a PA ranged from simple conditions such as an ear canal impacted with wax, to acute illnesses requiring immediate hospital admission via an emergency department and serious conditions requiring on-going care, such as leukaemia. The participants varied in their familiarity with the PA in their practice with 11 having consulted only once while the remaining 19 had consulted the PA previously. Of the latter, three had consulted very frequently with the PA in the management of an on-going condition.

| Thematic analysis
Four interlinking themes were identified as follows:

Patient willingness to see a PA again.
Each of these themes is described and exemplified below with quotes from the transcripts.

| Variation in understanding of the role of physician associates
The participants described the PA role in ways that varied widely.
We grouped participants' understandings into three categories: "certain and accurate," "certain and inaccurate" and "uncertain." The first two groups expressed their understanding of who they had seen with clarity, although their understanding may not have been accurate.
The first group was certain they understood the role of the PA and expressed this understanding accurately in terms of it being a close relationship to doctors, but correctly realizing that it was a different role, one which meant they had a recognized education but could not do everything a doctor could do. For example: My understanding would be somebody who's less qualified than a doctor but is able to deal with the sort of more routine things like earache I guess would be a good example of it.

(Participant 15)
Participants such as these recounted well-developed strategies within the practice of informing patients about the PA role, for example leaflets at reception and information given by the PA as soon as they entered the consulting room: The second of these groups was also confident in their perception of the role of the PA, but was inaccurate. They framed their description of the PA as being closely related to a doctor, for example understanding the PA as someone in training, "almost an apprentice" (Participant 17), or as a qualified doctor from another country who is simply unable to prescribe: Basically, as I understand it, they're basically a trained physician or trained doctor, but there's just a few things that they can't carry out, like signing the prescriptions and things like that, yeah.

(Participant 28)
These participants were therefore clear that there were differences between PAs and the doctors who were their GPs, and were aware of potential reasons for these but were not aware that the PA role was not in fact that of a doctor.
The third group was uncertain about the PA role. Of concern were those who had felt confident that they had seen a GP at the time of the consultation but had learned that they had seen a PA as a result Analysis of the participants' accounts therefore indicated that variability in understanding of the PA role was linked to the provision of information by the practice staff and by the PA, as well as to whether this was the first time they had seen the PA or had an on-going relationship with them.
The analysis of this theme then leads to the interlinked issue of trust and confidence in the physician associate and the general practice in which they were located.

| Trust and confidence in the physician associate
Participants were generally positive about trust and confidence in the physician associate and the consultation although some were more cautious or contingent. Trust and confidence appeared to be both influences on and influenced by the PA consultation through an interplay of health system (that is the NHS), their general practice and individual consultation level factors.
It was evident that confidence and trust were conferred on the PA Analysis of the issue of trust and confidence therefore highlights mixed, sometimes conflicting, experiences, apparently influenced by prior as well as "on the day" experience.

| Comparisons with a GP consultation
Participants were not specifically asked to compare their consultation with a PA to that of a GP but many did so in explaining their experiences in terms of what they usually received at their practice. Most participants perceived that their consultation with a PA was either no different from or was very similar to a consultation with a GP. They described being asked the same questions and given the same types of examination and investigations, as they considered they would have received from a doctor:

| Willingness to see a physician assistant again
The majority of participants reported that they were not offered a choice of whether they saw a PA or GP when they booked their same day/urgent appointment. For the small number of participants who described having actively sought an appointment with a PA, the reasons included a shorter waiting time to see a PA, dissatisfaction with prior appointments with GPs and trust in the PA based on previous contact.
Many participants expressed their willingness to see a PA in future consultations for any condition, while others expressed a willingness to return to consult a PA as conditional on the problem.

| DISCUSSION
Our findings presented differing patient experiences of consultations with PAs, although most were presented positively. Participants in general were unworried about the GP's task being substituted by a PA who appeared to act similarly to a GP, and who inspired high trust and confidence. However, participants were displeased if the role was not explained to them, feeling deceived by their practice and the PA.
Many felt that the PA was competent to perform a GP's role, but were sometimes frustrated by the restrictions around the role, particularly the inability to prescribe. Willingness to see the PA again was differentiated by presenting condition, as well as by experience and views on continuity of care.
This article has presented greater depth of understanding of the patient's perspective, as to the experience of consulting with a new type of health practitioner, a PA, who was substituting for a doctor in general practice. While the findings were broadly reflective of the larger survey's results, 29 the qualitative findings extend the knowledge available to those interested in the development and changes in skill mixes in providing same day or urgently requested primary care consultations -traditionally provided by doctors. 32 Interlinking influences on and impacts of patients' experiences have been identified which we present as a theoretical model illustrated in Figure 1 and discussed below.
The diversity of patients' understandings of the professional role of the PA ranging from "certain and accurate," through "uncertain" to "certain and inaccurate," has been identified in other UK studies of substitution by nurse practitioners for GPs 33,34 and also in primary care dental services in the substitution of dentists by dental therapists. 35 It was evident from the participants' accounts that the different forms of information used by general practices to explain the role had only been partially successful in ensuring that patients understood the nature of the physician associate substituting for the doctor. The absence of prior warning and explanation created situations in which confidence in the clinical care from the PA and in the general practice as a whole was at risk. Confidence and trust are linked concepts. 36 In health care characterized "by uncertainty and an element of risk regarding the competence and intentions of the practitioner on whom the patient in reliant" 37 (p2), trust is considered to be crucial. It was evident that where patient confidence in the PA, was apparent, it derived from the public health system, noted in one other substitution study, 35 but primarily from the general practice itself, as well as from the actions of the PAs themselves. Development of trust in nurse substitutes for doctors, through actual consultations, has been noted before. 33,38 We see a close relation to the model of Rowe and Calnan, 39  which has resulted in repeat consultations and more time being spent by patients in more visits to the general practice. 33,38 In the UK, a parliamentary Health Select Committee report has recommended that physician associates should be included in state regulatory processes as a matter of urgency, 40  This study was limited in that the volunteer participants were selfselecting, rather than purposively selected; however, they represented diversity in their characteristics and experiences of PA consultations.
Our practices and PAs were also volunteers and were small in number; F I G U R E 1 A representation of the interlinking influences on and impacts of patients' experiences of a physician associate in general practice in England we do not claim that these findings are generalizable, but the numbers of PAs in primary care are currently small and we achieved a range of practices. We chose to use telephone methods to overcome logistical problems which added to the immediacy following the consultation.
We are aware that, while there are suggestions from some studies that telephone interviews can yield lower quality in terms of missed reporting 43 , with the interviewer having no visual cues, 44 others conclude that the same amount and quality of data can be gathered in telephone and face-to-face interviews. 45