Nurse prescribing roles in acute care: an evaluative case study

Authors


K. Jones: e-mail: kathryn.jones@london.nhs.uk

Abstract

jones k., edwards m. & while a. (2010) Nurse prescribing roles in acute care: an evaluative case study. Journal of Advanced Nursing67(1), 117–126.

Abstract

Aim.  This paper is a report of an evaluation of the implementation of nurse prescribing in an acute care hospital in England.

Background.  At the time of the study, evaluation of nurse prescribing had taken place in community settings, but little was known about its impact and effectiveness in acute care. Although nurse prescribing has permitted doctor–nurse substitution in acute episodic care, some doctors have expressed concerns about patient safety in relation to nurse prescribing.

Methods.  A mixed methods single-case study was conducted in 2005–06, using purposive sampling. Semi-structured interviews were carried out with 18 hospital staff, non-participant observation of two nurses and two doctors undertaking 52 patient-prescriber consultations with 47 patients, and a questionnaire survey with 122 patients (response rate 61%: n = 74).

Results.  Nurse prescribing was found to benefit patients through service delivery improvement and using staff skills differently. Nurse prescribers and their colleagues were positive about role and service changes and their impact on patient care. No differences were found between the ways in which nurses and doctors performed prescribing roles, but there was a statistically significant difference between the medication-related information satisfaction ratings of patients who had seen a nurse prescriber, compared to those seen by a doctor.

Conclusion.  Nurses and doctors were found to provide equivalent care. Shared vision, local champions, action learning and peer support were the enabling factors that helped to embed the new prescribing roles within the study site.

What is already known about this topic

  •  Nurse prescribing works well in some primary care settings.
  •  Nurse prescribing has permitted doctor–nurse substitution in acute episodic care.
  •  Some doctors have expressed concerns about patient safety and nurse prescribing.

What this paper adds

  •  Nurses and doctors provided equivalent care, but patients who had seen a nurse reported higher satisfaction ratings of medicine related information.
  •  Patients across ethnic groups reported similar views about their prescribing experience and about their medicines.
  •  Shared vision, local champions, action learning, team, peer and buddy support were key to successful implementation of nurse prescribing roles.

Implications for practice and/or policy

  •  Employers should ensure that supportive and operational infrastructures are in place to underpin new prescribing practices.
  •  Future research is needed to measure the impact that nurse prescribing may have on improving patient outcomes.

Introduction

Nursing work is changing to meet the needs of healthcare systems as they adjust to the rising prevalence of chronic disease and growing numbers of older people within new cost constraints (WHO 2006). Escalating healthcare costs due to new technologies, new drugs and increasing public expectations, together with population changes, are drivers to deliver health care with improved efficiency and cost-effectiveness (Buchan & Calman 2004). Prescribing by nurses reflects the desire to maximize the capacity of the nursing workforce to improve both the speed of access to and the quality of care (Department of Health 2006) and has been implemented in a number of countries, including Botswana, South Africa, Sweden, Australia, New Zealand, the United States of America (USA) and Canada (Calman & Buchan 2004). In the USA, it is associated with advanced nurse practitioners across 50 states, where it is reported to improve patient outcomes, reduce healthcare costs and be well-received by patients (Brooten et al. 2002).

Background

Prescribing by community nurses in the United Kingdom (UK), which began in the early 1990s, has been extended to suitably qualified nurses in all settings (Department of Health 2002). While independent prescribers assume sole responsibility for the prescriptions that they write, including single accountability for patient assessment and any clinical decisions made, supplementary prescribers share responsibility for their prescribing with an independent prescriber, most frequently a doctor (Department of Health 2006). Latter et al. (2005), in a national evaluation of independent nurse prescribers (n = 246) in England, found that the majority of nurse prescribers felt strongly that their prescribing had had a positive impact on both patient care and patients’ access to their medicines, and had also enabled staff to use their skills better. Although a number of professional concerns were initially raised about the preparation of nurse prescribers and their competence to act (Avery & Pringle 2005, Chief Medical Officer 2005, Connelly 2005, Crown & Miller 2005, Siriwardena 2006), a national evaluation of supplementary prescribing confirmed its safety (Bissell et al. 2008). Indeed, non-medical prescribing has been shown to improve clinical decision-making across healthcare teams (Bradley & Nolan 2007) and enable reassessment of roles within multi-professional teams (Avery et al. 2007).

At the time of the study reported here, evaluation of nurse prescribing in the UK had focused on primary care (Luker et al. 1997, Latter & Courtenay 2004), with only two published evaluations in acute care settings (James 2004, Latter et al. 2005). Three studies had shown the positive views of patients, prescribers and their colleagues about nurse prescribing (Luker et al. 1997, James 2004, Latter et al. 2005), with only one study focusing solely on the views of patients (Brooks et al. 2001). The majority of the studies, both in primary and secondary care, had only reported nurse prescribers’ views of their own prescribing roles (Rodden 2001, Luker & McHugh 2002, Lewis-Evans & Jester 2004, While & Biggs 2004) to the neglect of other stakeholders. In the light of the existing research there was a need for further exploration of the nurse prescriber role, in particular an understanding of the settings in which the nurse prescribers worked, the differences that their new prescribing roles made to established ways of team working, and the views of nurse peers and healthcare colleagues across varied practice settings. Further exploration of the differences in the roles of medical and nurse prescribers was also needed to understand better the ways in which each prescriber worked, and the potential impact on patient outcomes, including satisfaction.

The study

Aim

The aim of the study was to evaluate the implementation of nurse prescribing in an acute care hospital setting in England.

Design

An embedded single-case design with three units of analysis (Yin 2003) was used (see Figure 1). The case study method was selected so that a clear set of propositions about the evolving phenomenon of nurse prescribing could be tested in a new context of new roles and ways of working in acute care (after Yin 2003). The units of analysis were chosen as examples of where the phenomenon could be studied at an operational level embedded within the case, and were used both as sources of data collection and units of analysis. We used multiple data collection methods, including semi-structured interviews, non-participant observation of patient-prescriber consultations and a patient survey using validated rating scales, to investigate the experiences of multiple stakeholders in relation to the implementation of nurse prescribing. The triangulation of data from multiple data sources can confirm or refute the evidence found (Hutchinson 1990). The main stakeholders were patients, prescribers and their colleagues (as recommended by Brooks et al. 2001, Latter & Courtenay 2004, Latter et al. 2005), and other hospital staff who held senior strategic positions. Propositions derived from practice knowledge were developed through discussions with staff and from an analysis of policy and professional literature, and were used to direct the study (see Table 1). An element within the case was a quasi-experiment to test the null hypotheses that there were no differences between the roles of medical and nurse prescribers.

Figure 1.

 The selected case.

Table 1.   Propositions derived from practice knowledge, with links to relevant literature
PropositionEvidence for proposition
Nurses spend more time with patients than doctorsLuker et al. (1997), Horrocks et al. (2002), Buchan and Calman (2004), Latter et al. (2005), Laurant et al. (2005)
Nurses give patients more information about their medicines than doctorsLuker et al. (1997), Brooks et al. (2001), Rodden (2001)
Patients who have medicines prescribed by nurses are more knowledgeable about their medicines than those who have medicines prescribed by doctorsWilson-Barnett and Beech (1994)
Patients who have medicines prescribed by nurses are more compliant with their treatment than those who have medicines prescribed by doctorsBrown and Grimes (1995), Berry et al. (2006)
Patients who have medicines prescribed by nurses experience fewer side effects than those who have medicines prescribed by doctorsTaylor et al. (1997)
Prescribing is a risky activity irrespective of the professional background of the prescriberCrown and Miller (2005)

Setting

One acute care hospital in England was selected as a case study. The hospital was located in a densely populated metropolitan borough with an ethnically diverse population [in 2001, 27% of the local population were from black and minority ethnic (BME) groups and 44% were born outside the UK], had 630 beds and employed 3500 staff to deliver general and specialist services. There were nine nurses who had undertaken preparation for prescribing, of whom seven (78%) were actively prescribing.

Participants

Three clinical departments were purposively selected as units of analysis because the doctors and nurses were early implementers of nurse prescribing and had been involved in the study design. These were the hypertension clinic, renal clinic and renal satellite unit. A purposive sample of staff was interviewed (n = 18) to explore the background and intended purpose of the new roles and the experiences of the nurse prescribers and their teams. Staff members were selected because they were prescribers (n = 3), mentors or colleagues (n = 7), or senior hospital staff (n = 8) who had influenced the implementation of new roles and ways of working within the organization, including the Directors of Medicine, Nursing, Human Resources and the Chief Pharmacist. The sample was not extended further because data saturation was achieved. Two nurses and two doctors from the hypertension and renal clinics were purposively selected for the observed opportunistic patient-prescriber consultations, through discussion with the lead clinicians for each service, as being exemplars of the new service delivery arrangements. A convenience sample of patients (n = 122) who attended the hypertension and renal clinics were invited to complete a questionnaire if they met the principal inclusion criteria of understanding an oral explanation of the study in English and were able to complete the questionnaire in English. Statistical advice was sought about the patient sample. Although the sampling strategy yielded a non-probability sample, the participants were chosen because of the likelihood that they would give information relevant to the study aim and would be most likely to confirm or refute the propositions held (Flyvbjerg 2001).

Data collection

Data were collected from July 2005 to September 2006 using semi-structured interviews with staff (n = 18) (July 2005–September 2006), structured non-participant observation of doctors (n = 2) and nurses (n = 2) conducting patient consultations (n = 52) (November 2005–January 2006), and a patient survey (n = 122) (November 2005–April 2006). In the semi-structured interviews we investigated both the background and intended purpose of the nurse prescribing roles and the experiences of the prescribers and of their colleagues, and were guided by questions formulated from the policy literature. Nurse prescribers were each interviewed on two occasions, once at the beginning of the study and then again 9 months later in a follow-up interview, when their colleagues were also interviewed. Senior staff were interviewed at the beginning of the study. Non-participant observation and a patient survey were used to investigate the null hypothesis. A structured observation sheet (15 items) was devised to assess prescriber competence (demonstrated/not demonstrated), in particular the ability to manage the patients’ medicine needs, based on a tool that had been used in a national evaluation (Latter et al. 2005) and an Observed Structured Clinical Examination (King’s College London 2004) tool in frequent use. The researcher (KJ) was an established practitioner who was familiar with structured observation for the assessment of clinical practice competence. Consultations were observed with patient consent at opportunistically selected clinics. Dual observation of consultations was not undertaken because it would have compromised the professional consultations, and this approach precluded an estimate of inter-rater reliability. The patient data were collected using a structured questionnaire (40 items) which included validated rating scales to assess respondents’ beliefs about their medicines (10 items) [Beliefs about Medicines Questionnaire (BMQ): Horne et al. 1999] and their satisfaction with the information they received from the prescriber (17 items) [Satisfaction with Information about Medicines Scale (SIMS): Horne et al. 2001].

Ethical considerations

The study was approved by the appropriate research ethics and governance committees. Four potential ethical issues were considered during the design and development of the study: the relationship between the researcher and the prescribers; a risk that the findings might suggest that nurse prescribers were either ‘not as good as’ or ‘better at’ prescribing than medical prescribers; a risk that patient consultations might be compromised by the presence of a researcher; and a risk that patients might not like the service provided. Participants in all components of the study were given study information sheets and time to consider participation, with the assurance for patients that their participation was entirely voluntary and would not jeopardize their care or treatment in any way.

Data analysis

The qualitative data were analysed using Ritchie and Spencer’s (1994) data analysis framework, which is a deductive approach that enabled the systematic sifting, charting and organization of the data according to key issues and themes. Quantitative data were analysed using the Statistical Package for Social Sciences Version 14 (SPSS Inc., Chicago, IL, USA). Numerical data were summarized using descriptive statistics. Non-parametric techniques were used to analyse categorical data and those that were not normally distributed.

Validity and reliability

In accordance with recommendations for case study rigour (Yin 2003), the study was guided by propositions. The interview schedule was guided by a literature review to enhance content validity, and the observation schedule was developed from tools that had been tested and used previously to measure the performance of nurse prescribers (King’s College London 2004, Latter et al. 2005). The patient survey questionnaire included items taken from validated rating scales designed to assess patients’ attitudes to and knowledge about their medicines (Horne et al. 1999, 2001), and general satisfaction questions taken from a national outpatient questionnaire (Healthcare Commission 2005) that enabled service comparison against national norms published as part of health service monitoring. Although the potential for researcher bias is acknowledged, a consistent approach to the data was adopted, including member checking (Stake 1995) by all participants at two levels: verification of interview transcripts and credibility of the thematic analysis and interpretation. The approaches to data analysis and interpretation were regularly checked within academic supervision.

Results

The sample comprised 18 staff who participated in the interview: nurse prescribers (n = 3), their medical, nursing and pharmacy colleagues (n = 7) and senior hospital staff (n = 8) (response rate 100%); of two nurses and two doctors who were observed undertaking 52 patient-prescriber consultations with 47 patients; and 122 patients who completed the questionnaire (response rate 61%: n = 74; 51% white, 47% from BME groups).

Background and intended purpose of the prescribing roles

The staff who participated in interviews unanimously reported that the overall intention of nurse prescribing was for patient benefit and that it yielded improved timeliness of care, better patient outcomes, improved team-working, advancement of working, legitimization of practice and increased patient and staff satisfaction:

Today it was great. I saw a … new patient today, and his blood pressure was [states value] and he’d run out of his medication and it would have taken him 48 hours at least to go back to his GP and get more medication. He hadn’t had medication for a week. He was at risk. I knew he had brought his tablets, I knew what he needed to take and I was able to write him a prescription. (Nurse Prescriber 1)

Experiences of nurse prescribers and their teams

At the time of their initial interview the nurse who worked in the hypertension clinic had been prescribing for 18 months, while nurses from the renal clinic and renal satellite unit had been prescribing for 11 months. All three nurse prescribers considered that their prescribing practice had had a positive impact on patient care and team-working, and had also enabled them to make better use of their nursing skills:

I can’t split it up [the impact of nurse prescribing]. If I think about how patients are sent to me here…our patients are very complex...The doctors and I share caring for those patients totally, and the way they use me is that they will see a patient and decide on the way they are going to go with medication, which is their skill, and then they will send the patient to me on a few occasions to titrate medication, add medication. And they will have decided what format to follow, and I will follow that format and I may discharge patients from that. Because patients may come to me it allows the doctor to see more patients that are more complex (that’s what their role is) and I can – rather than sending them back to the GP, the patient stays under our care, and I look after them in that format until they are ready to go back to the GP. (Nurse Prescriber 1)

The themes identified in the data set were: increased confidence to prescribe, improved patient care, increased role fulfilment and successful implementation of nurse prescribing. For example, all three nurse prescribers reported that they felt confident prescribing, but acknowledged that their confidence had developed with time. They all identified doctor colleagues, nurse managers and peers as key to the successful implementation of their prescribing role, and the importance of existing, and supportive, team-working, to their prescribing practice. They viewed the support of senior nursing staff positively – specifically an action learning set that had been established to provide support for novice nurse prescribers:

We do have a culture of nurses extending roles [in the renal unit] so then the support has already been there. Then it really, I think really, was the support of the Trust – the action learning and being part of the Trust, because it’s all very well doing it in renal but you do feel very isolated. So it was very important to come outside of that and have this opportunity to talk around the table. (Nurse Prescriber 3)

I decided to do it and was supported by the doctors because of their research [trials] and getting nurses to increase drugs. They need nurses who are good at it… they knew it was a safe thing for me to do. (Nurse Prescriber 1)

The colleague sample also unanimously reported the benefits of nurse prescribing to patient care, which was reflected in improved timeliness of treatment, improved patient care, improved team-work, influence on colleagues and changed team workloads:

It’s been really positive. They [the patients] are not waiting around for the doctors to sign prescriptions or for [names nurse prescriber] to discuss [treatment] with the doctors. She’s able to do the whole consultation and so it’s very positive for patients. (Nurse Colleague 5)

[She] sees patients in an individual fashion. She is extremely experienced in what they need and when they need it and the patients therefore get the drugs quicker. They don’t have to wait around so long. (Doctor Colleague 1)

Differences in roles of medical and nurse prescribers

Findings from the observation data

Forty-seven patient-prescriber consultations were observed, of which two-thirds (n = 30) took place in the renal clinic and one-third (n = 17) in the hypertension clinic. Sixty per cent (n = 28) of the consultations were between a patient and a nurse, and 40% (n = 19) were between a patient and a doctor. No differences were found in prescribers’ approaches to their patients or in the ways in which they managed their medicines. No statistically significant difference was found between the length of the observed patient-prescriber consultation and the profession of the prescriber (Chi-square test). Also, no difference was found in the prescribing practice of the doctors and the nurses, including the number and types of items prescribed per patient although some difference was noted in the methods of prescription used, for example, new or repeat medicine, but the dataset was too small to test for statistical significance.

Findings from the survey data

Seventy-four patients returned completed questionnaires (res-ponse rate 61%), of whom 51% (n = 38) were white and 47% (n = 35) from BME groups: Black Caribbean (n = 12); Black African (n = 7); Asian (n = 8); Chinese (n = 4); Mixed (n = 3); and Black other (n = 1). Seventy-two per cent (n = 53) of the patients had attended the renal clinic and 28% (n = 21) had attended the hypertension clinic. Three-quarters (n = 57) had consulted a nurse and one quarter (n = 17) had consulted a doctor, but 12 more patients reported that they had consulted a doctor (n = 29) than had actually done so. Patients unanimously reported confidence and trust in their prescriber, irrespective of whom they thought that they had consulted.

The beliefs held by patients about their prescribed medicines were measured using the BMQ (Horne et al. 1999). Specific-necessity and specific-concern scores were calculated for each of the 69 (93%) patients who had fully completed the BMQ. In a possible range of 1–5, scores ranged from 2·2–5 for specific-necessity and 1–5 for specific-concern. The scores indicated beliefs about the necessity of medicines for maintaining or improving health and about the potential adverse effects of taking medicines. No statistically significant difference was found between BMQ scores and the profession of the prescriber (Chi-square test). For example, 44% (n = 24) of patients who had consulted a nurse attained a specific-necessity score of 5, as did 40% (n = 6) of those who had consulted a doctor; the majority (53%, n = 8) of those who had consulted a doctor scored 4 for specific-necessity, as did 43% (n = 23) of those who had consulted a nurse. These scores indicated strong beliefs in the value of taking medicines to maintain or improve health. The majority (74%, n = 40) of the patients who had consulted a nurse attained a specific-concern score of 2–3, as did 73% (n = 11) of those who had consulted a doctor. These scores indicated low to moderate concern about the adverse effects of taking medicines. Whilst no difference was found between specific-necessity scores and patients’ ethnic origin, some difference was noted relating to specific-concern scores. For example, most patients with higher overall specific-concern scores (indicating the greatest concerns about their medicines) were from BME groups (64%n = 9), but the data set was too small to test for statistical significance.

Patients’ satisfaction with the medication information received from their prescribers was measured using the SIMS (Horne et al. 2001). Total satisfaction rating scores were calculated for each of the 64 patients (86%) who had fully completed the SIMS tool, and were grouped by the profession of the prescriber (Table 2). Scores ranged from 2 to 17 (possible range 0–17), with higher scores indicating a higher degree of satisfaction. There was a statistically significant difference between the satisfaction ratings of patients who had seen a nurse and those who had seen a doctor, with 66% of those (n = 33) who had consulted a nurse reporting a satisfaction rating of 17, compared to 7% (n = 1) of those who had consulted a doctor (χ² = 15·22, d.f. = 1, P < 0·001). Comparison of mean total satisfaction rating scores across the professional groups was also statistically significant, with nurses scoring a higher mean rank (36·56) than doctors (18·0) (χ² = 12·82, d.f. = 1, P < 0·001) using the Kruskal–Wallis test. There was also a small difference when comparing the scores of patients from different ethnic groups; for example, more patients from BME groups (59%, n = 20) attained a score of 17 than white patients (41%, n = 14), but the data set was too small to test for statistical significance.

Table 2.   Patients’ total satisfaction rating scores grouped by profession of prescriber
ProfessionTotal satisfaction rating scoreTotal
1–1011–1617
Nurse (n = 50)6
(12%)
(60%)
11
(22%)
(68%)
33
(66%)
(97%)
50
(100%)
Doctor (n = 14)4
(29%)
(40%)
9
(64%)
(32%)
1
(7%)
(3%)
14
(100%)
Total10
(100%)
20
(100%)
34
(100%)
64
(100%)

Discussion

Study limitations

In this small study we used a case study design in one acute hospital as little was known about the impact and effectiveness of nurse prescribing in the acute care setting; therefore the findings must be considered tentative until a larger study is conducted. Although a purposive sampling yielded a non-probability sample, every attempt was made to enhance validity and reliability at all stages of the research process. The use of a single observer without inter-rater observation may be a weakness; however, there were ethical considerations which took priority, as recommended by the research ethics committee. The use of video-recording is an alternative means of data collection which could have been considered in a study where the consultation process was the focus of enquiry. Unfortunately the dataset was limited, with fewer observed patient-doctor (n = 19) than patient-nurse consultations (n = 28), and fewer questionnaires from patients who had seen a doctor (n = 17) than had seen a nurse (n = 57); however, the data were sufficient to explore the phenomenon of interest in the context of a case study.

Benefits of nurse prescribing

Whilst the findings add to the growing body of knowledge about nurse prescribing, they are distinctive from others to date because we explored nurse prescribing from an organizational perspective using data from multiple levels, including senior managers and patients from a range of ethnic groups. Nurse prescribing was reported to be of patient benefit, echoing the findings of others (Luker et al. 1997, Latter & Courtenay 2004, Latter et al. 2005). Similarly, the qualitative data themes corresponding to the drivers for non-medical prescribing identified previously in the policy literature, namely quicker and more efficient access to medicines for patients and better use of staff skills (Department of Health 2002, 2006). The nurse prescribers in this study reported competence and confidence in prescribing, and considered that their prescribing practice had had a positive impact on both patient care and team-working and had also enabled them to use their nursing skills better, as found by others (Luker et al. 1997, Latter et al. 2005, Courtenay et al. 2006, 2007a, 2007b, Bradley & Nolan 2007, Carey et al. 2007, Courtenay & Carey 2008, Stenner & Courtenay 2008a, 2008b).

Doctors, nurse managers, peers and supportive team-working were identified as key to the successful implementation of nurse prescribing roles as reported by others (Latter et al. 2005, Courtenay et al. 2006, Bradley & Nolan 2007, Stenner & Courtenay 2008a, 2008b). In this study, nurse prescribers and their colleagues also made reference to the ‘fit’ of the new roles into existing team structures as helpful to implementation. For example, the hypertension and renal teams had reviewed uni-professional roles and joint ways of working in the light of the need to manage increased caseloads. However, it was the nurses themselves who were considered by managers, peers and colleagues to be crucial to the project’s success, through their motivation, enthusiasm and drive to succeed.

In contrast to Brooks et al. (2001) and Latter et al. (2005), we surveyed the views of patients from a variety of ethnic groups. All reported similar views about their prescribing experience and their medicines, irrespective of their ethnic background. No other study to date has explored the prescribing experience of patients from different ethnic groups. In fact, apart from an evaluation of the impact of prescribing by a diabetes nurse specialist (Carey et al. 2008), no other recent studies have included patients in their samples. This is both an omission and an opportunity for future research, given that modern healthcare systems are increasingly focusing on patient choice and improving the user experience (Department of Health 2008a).

Nurses and doctors were found to deliver equivalent care, but patients who reported the highest satisfaction ratings of medicine-related information had seen a nurse. A systematic review of the research into doctor-nurse substitution in primary care has shown that the quality of care is similar for patients seen by nurses as compared to those seen by doctors, with no appreciable differences in health outcomes, although higher satisfaction is reported by some patients (Horrocks et al. 2002). Systematic reviews of the research literature have also shown no appreciable differences in the care delivered by doctors and nurses, although nurses spend longer with patients (Horrocks et al. 2002, Buchan & Calman 2004, Laurant et al. 2005). However, in the present study the majority of the patients with the longest consultation times were seen by a doctor, which may partly be explained by doctors seeing patients with more complex morbidity and therefore needing longer consultations. Thus, patients who reported higher satisfaction ratings had not spent longer with their prescribers. This is an important finding for healthcare professionals in England, where policy directives focus on service improvements through changes in delivery and role redesign (Department of Health 2008a, 2008b). The process of training team members to prescribe has also been shown to enable reassessment of roles (Avery et al. 2007) and improve clinical decision-making across healthcare teams (Bradley & Nolan 2007).

A synthesis of the findings (after Lewin 1951) offers a potential model of the driving and resisting forces identified from the dataset which have an impact on implementation of new nurse prescribing roles within a chosen site (see Figure 2). These driving forces included viewing workforce changes as an opportunity, improved patient care, better medicine management, effective sponsorship and support of change, team engagement, advanced practice and expert care. The restraining forces were evolving infrastructure and emerging policy. Thus, we recommend that employers ensure that supportive and operational infrastructures are in place to underpin new prescribing practices if a major change is to be successfully implemented. Most importantly, four variables were identified as actively enabling the implementation of nurse prescribing roles: shared vision, local championship, action learning and team, peer and buddy support.

Figure 2.

 A force field analysis for the implementation of nurse prescribing (Strength of line indicates strength of supporting evidence derived from the data set. Dotted lines indicate forces arising as a result of the implementation of the new service model).

Conclusion

In the light of growing evidence of the competency of nurse prescribers, it is now timely to focus less on behavioural and affective measures of prescribing performance and more on the impact that the role may have on enhancing the quality and safety of patient care, such as identifying those interventions that improve patient safety by reducing prescribing errors. In this way, the benefits or otherwise of nurse prescribing roles can be more comprehensively assessed, using clearly defined performance indicators and patient outcomes and findings compared with those derived from preliminary reports on nurse prescribing in terms of ‘equivalency to’ or ‘substitution for’ doctors only.

Acknowledgements

With acknowledgement to Jill Bunker, Wendy Brown and Jen McDermott, the nurse prescribers who participated in this study.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of interest

No conflict of interest has been declared by the authors.

Author contributions

KJ was responsible for the study conception and design. KJ performed the data collection. KJ performed the data analysis. KJ was responsible for the drafting of the manuscript. ME and AW made critical revisions to the paper for important intellectual content. ME and AW provided statistical expertise. ME and AW supervised the study.

Ancillary