1 INTRODUCTION

Face-to-face primary care consultations grew by more than 15% between 2010/2011 and 2014/2015, partly due to the ageing population (Majeed, 2014; The King's Fund, 2016). Alongside this, employment of full-time equivalent general practitioners (GPs) has decreased (NHS Digital, 2018). There have been efforts to increase the number of advanced practitioners (APs), partly to fill shortfalls, but also to allow patients early access to specialists (NHS England, 2016). APs are healthcare professionals (HCPs) working at a higher level than initial registration and they should work in, lead and manage a multidisciplinary team; critically address their learning needs; and engage in research (Health Education England [HEE], 2017). Examples of APs include first contact physiotherapists (FCPs), clinical pharmacists, primary care practitioners and the more well-established nurse practitioner (NP) role (HEE and NHS England, 2018; NHS England, 2016).

Ascertaining the patient perspective is vital; if an intervention is considered acceptable, patient adherence to treatment and improved clinical outcomes are more likely (Hommel et al., 2013). Patient satisfaction is defined around patient beliefs and expectations being met, whereas acceptability is a multifaceted construct (Sekhon et al., 2017). This realist synthesis will explore patient ‘views’, encapsulating multiple measurements of the patient perspective.

1.1 Aim

Explore the literature on patient views of the AP role in primary care to determine the factors that influence acceptability.

1.2 Objectives (1)

Identify literature relevant to patient acceptability of the AP role.

(2)

Interrogate relevant literature using realist theory.

(3)

Establish hypotheses on what makes the AP role acceptable/unacceptable to patients.

(4)

Establish the underlying contexts, mechanisms and outcomes of these hypotheses.

2 METHODS

Realist synthesis was adopted, a method which follows the ontology and epistemology of empirical realism and can be undertaken prior to a realist evaluation. Realism is concerned with ‘what works for whom, how and under what circumstances’ (Pawson et al., 2005, p. 32). Realist syntheses are suited to complex interventions, such as health care (Rycroft-Malone et al., 2012).

Realists make causal links between the ‘context’, ‘mechanism’ and ‘outcome’, known collectively as ‘context-mechanism-outcome (CMO)’ configuration’. A programme (the AP) operates within a context which ‘triggers’ the mechanism that creates an outcome (Wong et al., 2016). In realist syntheses, CMOs are identified from relevant literature and then analysed to form programme theories whilst ensuring stakeholder involvement throughout (Pawson et al., 2005).

3 DEFINING THE SCOPE

A broad search of sources formed initial ideas on how the AP role works; these shaped the theory areas (TAs) that acted as the framework for the development of hypotheses (Rycroft-Malone et al., 2012). Two FCPs, a research associate (involved in FCP research) and a patient research partner (referred to collectively as the ‘team’) met with the researcher (LM) to discuss the TAs, alter as needed or create new TAs.

Seven initial TAs were agreed which formed the TA framework:

TA1—Patient's prior experience of condition management

TA2—Patient's expectations of condition management

TA3—Communication

TA4—Continuity of the individual practitioner

TA5—Practitioner's scope of practice

TA6—Accessibility

TA7—Promoting the role

4 SEARCHING AND APPRAISING THE LITERATURE

An overview of the process for searching and appraising the literature is outlined in Figure 1.

Searching and appraising the literature. Total number of studies included to read at title level after duplicates removed n = 2274. Read at full-text level n = 65, including 14 from snowballing

A realist synthesis utilises purposive searching of proposed theories and iterative searches, as programme understanding grows (Pawson et al., 2005). Databases were searched between 30 May 2017 and the 26 October 2017 and included the Allied and Complementary Medicine Database; CINHAL Plus; Medline; PsycARTICLES; PyscINFO; PEDro and the CSP's Evidence and Knowledge Discovery Search Service. Searching ceased once it stopped adding to the understanding of the intervention (Pawson et al., 2005). Seven different search strategies were adopted, with terms specific to each theory area and its hypotheses.

Sources were not assessed based on the study design or quality; therefore, grey literature was included (Rycroft-Malone et al., 2012) (see Table 1 for inclusion/exclusion criteria).

TABLE 1. Inclusion/exclusion criteria Abbreviation: AP, advanced practitioners.

The search terms, database, number of hits (titles) and duplicates removed by the database were recorded.

5 DATA EXTRACTION AND APPRAISAL

The data extraction sheets (n = 7) collated information on each TA, with questions regarding CMOs, and were piloted by two researchers (LM and JP) (Rycroft-Malone et al., 2012). Included literature was reviewed and data extraction involved transfer of phrases, sentences or sections relevant to the TA.

6 ANALYSIS

The analysis was conducted in five steps. (1) Data extraction and analysis was presented simultaneously in a table, alongside the information from the following steps. (2) Themes were recorded by the lead researcher (LM) and validated by all team members. (3) Formulation of chains of inference—the connections across themes that create an overarching theme (Rycroft-Malone et al., 2012). (4) Chains of inference were connected through identifying studies with the same chains and recording study number(s). (5) Hypotheses are synthesised statements of findings (Rycroft-Malone et al., 2012). They were formed from the accumulative picture of CMOs subsequent to data extraction of all studies. To form theory, available evidence, hunches, common-sense and team discussions were utilised (The RAMESES II Project, 2017). An example data extraction sheet is provided in Appendix 1.

7 RESULTS

Seven theories were identified from the scoping review, expanded upon and redefined through a systematic review of each TA, resulting in 19 hypotheses related to the patient acceptability of the AP role.

2274 articles were read at title/abstract level and 65 at full-text level, including 14 from snowballing. Thirty-seven articles were included in the review; 5 regarding the physiotherapy FCP role, and 32 studies were nursing roles, Health Visitors, Physician Assistants or Pharmacist Independent Prescribers. A new TA ‘professional hierarchy’ emerged at the data extraction phase through reading the literature related to the other seven TAs (see Figure 2). The narrative for the hypotheses (Hs) is presented under the relevant context and Table 2 provides further CMO evidence.

TABLE 2. Context–mechanism–outcomes Theory area Title Context Mechanism and outcome Evidence Experience 1. AP consultation can lead to the equivalent type of outcomes/treatments as a GP consultation. Experience of previous GP consultations Resource M: AP offered alternatives to prescriptions. Response M: Patients evaluate AP role through comparison with the GP. (Resell et al., 2007; Bergman et al., 2013; Chartered Society of Physiotherapy, 2017) Expectation certain answers to questions/recognition serious illness/examination/onward referral/prescription. O: Dissatisfaction if expectation not met OR increased satisfaction for receiving an alternative to a prescription. Expectations 2. Patient perceptions of 'serious' conditions affects the acceptability of the AP consultation. Resource M: AP does not diagnose a serious condition. Response M: Patient expected confirmation of their belief of a serious condition due to preconceived ideas from GP consultations. Unintended O: Patient dissatisfied with AP as their expectation was not confirmed.

(Barratt, 2016; Caldow et al., 2006; Halcomb et al., 2013; Luker et al., 1998; Mahomed et al., 2012; Maul et al., 2015; Myers et al., 1997; Parker et al., 2012; The EROS Project Team, 1999; Young et al., 2016)

Experience 3. Patient perceptions of GPs formed from previous GP consultations will influence the patient acceptability of the AP role. Resource M: AP provided exercise advice. Response M: Patient expected AP to prescribe as their GP had previously. (Resell et al., 2007; Gerard et al., 2014) O: Increased patient satisfaction from exercises. Communication 4. The AP role is more acceptable to patients when the AP has an informal discussion with the patient. Resource M: AP friendly and consultation more conversational than GP. Response M: Patient able to build a rapport with the AP. O: Patients more likely to ask AP questions. (Halcomb et al., 2013; Mahomed et al., 2012; Maul et al., 2015; Myers et al., 2015; The EROS Project Team, 1999; Perry et al., 2005; Phillips et al., 1998; Williams & Jones, 2006) Accessibility 5. Longer consultation lengths increase patient acceptability of the AP role. Resource M: AP consultation longer than GP consultation. AP answered all patient's questions. Response M: Patients perceived APs had more time for them than GPs and that they explained things fully. O: Felt more enabled to manage own health. (Barratt, 2016; Halcomb et al., 2013; Luker et al., 1998; Young et al., 2016; Williams & Jones, 2006; Redsell et al., 2006; Dhalivaal, 2011; Reveley, 1998; Wynne, 2016; Webster, 2008) Continuity 6. Having familiarity with the practitioner in the consultation increases patient acceptability of the AP role. Resource M: Due to practitioner continuity, the AP knew the patient's name/history. Response M: Patient expectation to build a long-term relationship with the AP which they were unable to do with GP. O: More confident in AP. (Bergman et al., 2013; Barratt, 2016; Halcomb et al., 2013; Luker et al., 1998; Mahomed et al., 2012; Phillips et al., 1998; Williams & Jones, 2006; Redsell et al., 2006; Desborough et al., 2016; Edwall & Danielson, 2008; Fortin et al., 2010; Brooks et al., 2001; Chapple et al., 2000) Promoting the role 7. A greater understanding of the AP role increases patient acceptability of the role. Latent resource M: Limited promotion of the AP role to patients. (Barratt et al., 2016; Caldow et al., 2006; Maul et al., 2015; Williams & Jones, 2006; Reveley, 1998; Chapple et al., 2000; Wasylkiw et al., 2009; Baldwin et al., 1996) Unintended response M: Inappropriate self-referral. O: Patients have the wrong understanding of the AP. Inappropriate referrals lead to over-demand OR appropriate patients not accessing the role. Communication 8. The role is more acceptable to patients when AP's are person-centred in their consultation style. Patients want to be able to make decisions in their care Resource M: AP explained information in an accessible way and provided holistic advice. Offered alternatives to medications. Response M: Patient felt APs used language they could understand. Felt they were receiving person-centred care. Patients felt empowered to manage their condition. (Barratt, 2016; Caldow et al., 2006; Halcomb et al., 2013; Gerard et al., 2014; Redsell et al., 2006; Dhalivaal, 2011; Chapple et al., 2000; Kernick et al., 1999; Shum et al., 2000) O: Patients preferred being educated by the AP than the GP. Patients able to self-manage. 9. The AP role is more acceptable to patients when the AP demonstrates a high-level of knowledge. Patients want to understand clinicians' clinical reasoning Resource M: AP thorough in their assessment and information provision. Response M: Patient perceived AP as knowledgeable. O: Patient trusting of AP and more satisfied with the outcome of their consultation. (Barratt, 2016; Halcomb et al., 2013; Luker et al., 1998; The EROS Project Team, 1999; Perry et al., 2005; Dhalivaal, 2011; Reveley, 1998; Webster et al., 2008; Brooks et al., 2001; Langer, 1995; Tinelli et al., 2013) Experience 10. Limited prior experience of an AP decreases patient acceptability of the role. Experience of an AP Resource M: APs have specialist capabilities. (Gerard et al., 2014; Chapple et al., 2000; Wasylkiw et al., 2009; Baldwin et al., 1996) Response M: Patients feel uncomfortable with the AP's specialist capabilities. Unintended O: Patients less likely to access AP for prescriptions. Scope of practice 11. Role more acceptable if AP offers a service that is equivalent to the GP consultation. Resource M: AP able to carry out medical investigations. Unintended response M: Uncomfortable with AP's scope. Unintended O: Less likely to access an AP if requiring a skill that is an extension of their scope. Inactive O: Increased enablement of patient to manage their own condition when accessing an AP with a greater scope of practice. (Parker et al., 2012; Gerard et al., 2014; Desborough et al., 2016; Wasylkiw et al., 2009; Baldwin et al., 1996) Experience 12. Previous experience of a prescribing AP increases patient acceptability of a prescribing AP in primary care. Resource M: AP able to prescribe. Response M: Patients more accepting of AP prescribing due to previous experience. O: Patient accesses AP for prescription. Patient more enabled to manage their own health when the AP has a greater scope of practice. (Gerard et al., 2014; Desborough et al., 2016) Expectations 13. Patients less accepting of the role if prescriptions are not checked by the GP. Patient perceives they have a ‘serious’ condition Resource M: AP able to prescribe. Unintended response M: Patients expected APs to prescribe for ‘simple problems’ OR patients expect AP to discuss prescriptions with GPs to reduce risk. O: Reduced acceptability of AP prescribing unless they access the GP. (Bergman et al., 2013; Caldow et al., 2006; Brooks et al., 2001) 14. Lack of patient choice decreases patient acceptability of the FCP role. Resource M: Patient choice of GP or AP access. Response M: Patients selective on who they want to diagnose their perceived 'serious' condition. Patients want to retain the choice to access their GP. O: Patient expectation that they should see the GP if their conditions perceived as ‘serious’. Patients selective on what conditions they access GP for. (Halcomb et al., 2013; Mahomed et al., 2012; Maul et al., 2015; Fortin et al., 2010; Holdsworth et al., 2004) Accessibility 15. A decrease in waiting times for services increases patient acceptability of the AP role. Long wait for a GP appointment Resource M: Reduced wait for AP appointment. (Bergman et al., 2013; Halcomb et al., 2013; Luker et al., 1998; Myers et al., 1997; Young et al., 2016; Perry et al., 2005; Reveley, 1998; Edwall & Danielson, 2008; Brooks et al., 2001; Kernick et al., 1999; Shum et al., 2000; Langer et al., 1995; Heale & Pilon, 2012; Roblin et al., 2004) Response M: Patients felt calmer if they could access AP when required. O: Increased acceptability if reduced waiting times for AP. Patients seen earlier and reassured. ‘Freeing up’ GP appointments. 16. Patients find the role more acceptable if they expect that an engagement with FCP will provide indirect access to other services. Resource M: Follow-up appointment with AP. (Luker et al., 1998; Fortin et al., 2010) Unintended response M: Expected AP to be first appointment only, and GP follow-up to. Expected AP to be able to expedite access to a GP. O: Hesitancy with AP follow-up. 17. Increased acceptability of the role if the service is more convenient to the patient. Resource M: AP able to prescribe. (Bergman et al., 2013; Caldow et al., 2006; Dhalivaal 2011; Brooks et al., 2001; Tinelli et al., 2013; Heale & Pilon , 2012) Response M: Patients perceive AP as convenient as they can receive their prescriptions earlier, in a reduced number of appointments. O: Increased acceptability of the AP if they can prescribe. Prescription in one appointment. Promoting the role 18. Peer validation influences patient acceptability of the AP role. Patients sharing their healthcare experiences with others Resource M: Local press/patients telling others about the AP role. Website, multimedia methods, notice board and leaflets promoting the role. Barratt 2016; Maul et al., 2015; Baldwin et al., 1996) Response M: Increased patient awareness of the role through patient word-of-mouth only. O: Increased access of the AP (not quantified) Promoting the role 19. GP Practice staff validation increases patient acceptability of the AP role. The role of GP Practice staff in signposting Resource M: Receptionists promoting the role and highlighting earlier access/ability to see the same individual. GPs explain the AP role to patients. Response M: Patient aware of AP role. O: Increased patient understanding of the role. Unintended O: Over-demand of the service. (Halcomb 2013; Williams & Jones, 2006; Webster et al., 2008; Desborough et al., 2016; Fortin et al., 2010; Chapple et al., 2000; Cook et al., 2013) Hierarchy Novel 1. ‘professional hierarchy’ was a new theory area that was not in the initial framework. Therefore, this theory area was not actively searched for, literature was limited and therefore no hypotheses were formed. Hierarchy within the professions Resource M: AP demonstrates a high-level of knowledge. (Barratt, 2016; Redsell et al., 2006; Chapple et al., 2000) Response M: Presumption that the AP must have a lesser level of knowledge. Expect the AP to carry out traditional roles. Patient mistook AP to be a GP. O: Insufficient data. Abbreviations: AP, advanced practitioners; FCP, First Contact Physiotherapists; GP, general practices. 8 EXPERIENCE OF PREVIOUS GP CONSULTATIONS/APS

Patients formed expectations of AP consultation outcomes through comparison with GP outcomes (Baldwin et al., 1996; Gerard et al., 2014; Wasylkiw et al., 2009) (Hs1–3). Patients were more comfortable with a ‘friendly’ AP than with a GP (Barratt, 2016; Gerard et al., 2014; Mahommed et al., 2012; The EROS Project Team, 1999) and, consequently, more likely to ask their AP questions (Dhalivaal, 2011; Phillips et al., 1999; Redsell et al., 2006) (H4). Patients valued longer AP consultations (compared with GPs) (Halcomb et al., 2013; Luker et al., 2008; Reveley, 1998; Williams & Jones, 2006; Young et al., 2016) (H5) which increased their satisfaction (Desborough et al., 2016; Roblin et al., 2006). However, three studies postulated that efficient communication may have created the illusion of APs having more time (Barratt, 2016; Redsell et al., 2006; Shum et al., 2000). When patients had a long-term relationship with their GP, they expected to create a similar AP relationship (H6) (Fortin et al., 2000).

Patients wanted education on AP qualifications (Baldwin et al., 1996; Reveley, 1998), training (Caldow et al., 2006; Reveley, 1998), specialist skills (Caldow et al., 2006; ), and what the role can offer and how to access it (Chapple et al., 2000; Webster et al., 2008) (H7).

Patients valued the AP being person-centred in their approach (Barratt, 2016; Caldow et al., 2006; Dhalivaal, 2007; Luker et al., 1998); this was compared to their GP experience (Myers et al., 1997; Williams & Jones, 2006) and was particularly important for patients who wanted to be involved in decision-making (Dhalivaal, 2007) (H8). Patients were uncomfortable with APs' scope of practice if they had limited experience of it or an associated role (Baldwin et al., 1996); consequently, they were less likely to access an AP (Gerard et al., 2014; Wasylkiw et al., 2017) (Hs10–11).

9 PATIENTS WANT TO UNDERSTAND CLINICIANS' CLINICAL-REASONING/MAKE DECISIONS IN THEIR CARE

Patients wanted the AP to display their knowledge via a thorough assessment and clearly explained information (CSP, 2016; Dhalivaal, 2007; Redsell et al., 2007) (H9). This was associated with the ability for patients to make decisions regarding their care (Edwall & Danielson, 2008). Studies highlighted a patient desire for shared decision-making (Barratt, 2016; Mahomed et al., 2012; Young et al., 2016) (H8).

10 PATIENT PERCEIVES THEY HAVE A ‘SERIOUS’ CONDITION

Although patients were satisfied with—and often expected—APs to prescribe medications (Barratt, 2016; Bergman et al., 2013; Redsell et al., 2007), the expectation remained that they would discuss the prescription with a GP to reduce risk (Bergman et al., 2013) (H13).

Patients wanted to retain the choice of GP access, feeling that GPs had more in-depth knowledge and should diagnose what were considered ‘serious incidents’ (Halcomb et al., 2013). Serious incidents were predominantly related to the existing conditions (Maul et al., 2015) while patients would consult APs for common colds and infections (Barratt et al., 2016; Myers et al., 1997). Holdsworth and Webster (2004) found that self-referred patients were more likely to have had their condition for a shorter duration (H14).

11 LONG WAIT FOR A GP APPOINTMENT

Patient dissatisfaction with long GP waiting times increased the acceptability of AP consultations when waits were shorter (Bergman et al., 2013; Halcomb et al., 2013; Heale & Pilon,