Nurse practitioners and physician assistants working in ambulance care: A systematic review

Background: This review aims to describe the activities of nurse practitioners (NPs) and physician assistants (PAs) working in ambulance care, and the effect of these activities on patient outcomes, process of care, provider outcomes, and costs. Methods: PubMed, MEDLINE (EBSCO), EMBASE (OVID), Web of Science, the Cochrane Library (Cochrane Database of Systematic Review), CINAHL Plus, and the reference lists of the included articles were systematically searched in November 2019. All types of peer-reviewed designs on the three topics were included. Pairs of independent reviewers performed the selection process, the quality assessment, and the data extraction. Results: Four studies of moderate to poor quality were included. Activities in medical, communication and collaboration skills were found. The effects of these activities were found in process of care and resource use outcomes, focusing on non-conveyance rates, referral and consultation, on-scene time, or follow-up contact Conclusions: This review shows that there is limited evidence on activities of NPs and PAs in ambulance care. Results show that NPs and PAs in ambulance care perform activities that can be categorized into the Canadian Medical Education Directives for Specialists (CanMED) roles of Medical Expert, Communicator, and Collaborator. The effects of NPs and PAs are minimally reported in relation to process of care and resource use, focusing on non-conveyance rates, referral and consultation, on-scene time, or follow-up contact. No evidence on patient outcomes of the substitution of NPs and PAs in ambulance care exists. PROSPERO registration: CRD42017067505 (07/07/2017)


Background
Ambulance utilisation has increased in the Western world over the past 20 years, potentially compromising access, quality, safety, and patient outcomes 1,2 . Population ageing, changes in social support and accessibility, increasing community health awareness, patients presenting themselves with higher complexity and comorbidities, repeated ambulance care requests, and ambulance care request for primary healthcare problems have been described as associated factors for this increase 1,3-6 . The pressure to reduce costs and the potentially negative effects of this increase of ambulance utilisation have led to the redefinition of the roles of professionals in prehospital care 1,2 . With the impending rise in demand for health services, an effective utilization of the workforce is paramount to ensure highquality yet cost-effective health service delivery 7 . This can be done by optimising triage and ambulance allocation, but also by introducing other types of healthcare professionals in the ambulance domain. A possible solution to improve the balance between the increasing demand for care and the decreasing supply of medical healthcare workers is enhancing the role of allied healthcare workers, such as nurse practitioners (NPs) or physician assistants (PAs) 8  Several reviews about the implementation of NPs and PAs have been performed [9][10][11][12] . These reviews have revealed not only a higher quality of care but also an increase in patient satisfaction and that NPs and PAs have the potential to reduce doctors' workloads and direct health care costs. However, this research has been limited to long-term care facilities and primary health care; there currently is no evidence pertaining to what activities NPs or PAs in ambulance care perform and what the effects of these activities are.

Aims
Therefore, this review has two aims. First, to describe the activities of nurse practitioners and physician assistants working in ambulance care. Second to describe the effects of these activities on patient outcomes, process of care, provider outcomes, and costs.

Design
This study is a systematic literature review reported according to the steps of Cochrane Handbook 13 and reported to conform with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement. For background and an extensive method section, see the study protocol 14 .

Search strategy
The Cochrane Database for Systematic Reviews and PROS-PERO were inspected for similar reviews or protocols. No (pending) review was identified, so systematic searches were performed in PubMed, MEDLINE (EBSCO), EMBASE (OVID), Web of Science, the Cochrane Library (Cochrane Database of Systematic Reviews), and CINAHL Plus in November 2019.
Search strategies were developed to represent terms for ambulance care and NPs or PAs. Full search strategies per database are provided as extended data 15 .
Study selection procedure Searches were not restricted by year of publication. All types of peer-reviewed systematic reviews, and quantitative or qualitative designs in real clinical practice or simulation situations on NPs or PAs working in ambulance care for all kinds of patients were included. Conference abstracts, narrative reviews, editorials, personal communications, and unpublished studies were excluded.
Studies were included if they (a) described activities of professionals with a master's degree in ambulance care (NPs or PAs) and/or described the effects of the NP or PA on patient outcomes, process of care, provider outcomes and costs.
Due to the heterogeneity of the names that are used for the emergency medical service (EMS) professional worldwide 16 , we began with a broad search. Some terms covered a variety of different professionals; for example, the education level of the emergency care practitioner (ECP) can be that of a paramedic or a nurse. We explicitly searched for professionals with a master's degree and excluded studies where this was not present.
Two reviewers (RvV and ND) independently screened the title and abstract of each potentially relevant study. Differences between the reviewers were resolved through discussion. Couples of two independent reviewers (RvV, RE, ND, JL, and LV) screened the full texts of the remaining articles. In addition, two reviewers (RvV and RE) screened the reference lists of the included articles.

Quality assessment
To assess the quality of observational studies, we used a tool developed for evaluating primary research papers in a variety of fields 17 . Couples of independent researchers (RvV, RE, ND, JL, and LV) performed this assessment. Differences between two reviewers were resolved through discussion; in cases of doubt, a third reviewer from another couple made the final decision.
Data extraction, synthesis, and presentation Couples of independent researchers (RvV, RE, ND, JL, and LV) extracted the data. Due to the heterogeneity of study designs, settings, countries, care providers, interventions, and outcome measures, a meta-analysis was not possible; the results of this systematic review are therefore presented in tabular form.

Review statistics
The initial search identified 1283 unique records; 68 articles were included for full-text screening (see Figure 1), from which we included four articles for data extraction. A list of excluded articles and the reasons for their exclusion (n = 64) is provided as extended data 18 . Common reasons for exclusion were a non-master educational level, the lack of peer review, and the wrong setting (not prehospital ambulance care).
Study characteristics (Table 1) Designs of the included studies comprised of one cross sectional 19 , one retrospective observational 20 , one action research 21 , and one retrospective descriptive review 22 (Table 1). Two studies were performed in the UK, one in the Netherlands and one in the USA. All these studies extracted the data from ambulance run records or patient records. The focus of three of these studies 19,21,22 was primarily on ambulance care, where the retrospective observational study 20 had a broader perspective of home care, ambulance care, and emergency care. One study 19 compared the PA with a registered nurse (RN), two compared the NP with other EMS professionals (e.g., paramedics) 20,21 and one solely described the NP 22 without comparison.
Quality assessment (extended data 23 ) The cross sectional study 19 is of moderate quality due to the representativeness of results, and small population. The other three studies 20-22 are of poor quality. Table 2). Two articles reported on activities NPs or PAs perform in ambulance care 19,20 , these activities were related to medical skills, communication and collaboration. For medical skills, the usage of the SCEBS methodology and overall care, assessment, investigation, management, and quality of record registration were described 19,20 . For communication, the provision of medical advice and for collaboration referral to the ED or GP were described 19 .  Effects of activities (Table 3) None of the included studies reported on patient outcomes, care provider outcomes, or costs; the studies did report on process of care and resource use.

Activities by a NP or PA (
All four studies 19-22 reported on process of care outcomes. The outcomes used included the proportion of non-conveyance, the number of referrals in non-conveyance patients, the number of consultations, the length of on-scene treatment, the follow-up contact of non-conveyance patients, diagnostic measurements, adherence to guidelines and protocols and, number of performed interventions. One study reported on resource use.

Non-conveyance (n=3).
Three studies reported on non-conveyance 19,21,22 and showed non-conveyance rates ranging from 20% -50% for PAs. Non-conveyance rates for the NP were not described.

Referral and consultation (n=1).
One study 19 found that PAs refer 50% of their patients to another health care professional (e.g., a GP or an emergency department (ED)) while RNs referred 73%. Furthermore, PAs consulted other health care professionals (e.g., a GP, an emergency physician, or a medical specialist) significantly more often compared to RNs.

On-scene time (n=2).
One study found no significant difference between PAs and RNs regarding the length of on-scene treatment time 19 . Another study described 22 an average length of treatment time on scene of 21.47min, but made no comparison with other EMS professionals.

Follow-up contacts (n=1).
Follow-up contact after the completion of prehospital EMS care also indicated no significant differences between PAs and RNs 19 .

Discussion
This review aimed to describe which activities NPs or PAs deploy in ambulance care, and if there were effects on patient outcomes, process of care, provider outcomes, and costs.
The results indicate that little is known on the activities PAs and NPs deployed in ambulance care. This can be explained by the relatively young professions these professionals represent. The activities that were identified can be categorized using the Canadian Medical Education Directives for Specialists (CanMED) framework. The CanMEDS system is a widely used instrument to describe medical professionals activities and forms the basis of the education of NPs and PAs 24 . A competent professional seamlessly integrates all seven competencies CanMEDS roles 24 (Medical Expert (the integrating role), Communicator, Collaborator, Leader, Health Advocate, Scholar and Professional). However, the activities found in this review can be categorized into the medical professional, communicator and collaborator. This is remarkable because the full NP and PA profiles includes seven CanMEDS roles. There are several reasons why all seven roles are not reported on. First, it is possible that not all seven roles are applicable in ambulance care, or are not visible in the primary process of ambulance care. Also, PAs and NPs have only recently integrated into the ambulance care system, a clear job description or interpretation of their duties may be lacking. Developing a systematic description of the roles and competences of NPs or PAs in ambulance care would therefore be useful.
Although there are differences in education between NPs and PAs, there also seems to be a large degree of overlap in the tasks that NPs and PAs perform in practice 25 ; for instance, both professionals perform tasks that are part of the medical process, such as, drafting and evaluating treatment plans, and carrying out interventions 25 .
Results shows that little is known on the effects of the activities of NPs and PAs in ambulance care. Some effects found can be linked to process of care and resource use. We found no effects on patient outcomes or care provider outcomes. Reviews 12,25,26 in other health care settings revealed an increase in quality of care and patient satisfaction. Evidence in primary care 26 , elderly care 3 , and out-of-hours primary care 27 suggests that the substitution of NPs or PAs is feasible with at least the maintenance of quality and no increase in costs.
Although we have found no description on the effect on costs, Walsh et al. 21 assumed that the substitution of NPs could produce substantial savings for the EMS and relieve the burden on hard-pressed ambulance and ED. Bloemhoff et al. 19 recommended further exploration into the costs.
Further research is necessary to draw any conclusions on the effects of the substitution of NPs and PAs in ambulance care for multiple outcomes. This should be addressed by using the six dimensions of quality of healthcare: 1-effectiveness, 2-efficiency, 3-patient safety, 4-accessibility, 5-timeliness and 6-target population directed 26 . Measuring these outcomes within all phases of the ambulance process (from initial call, to handover or referral) will gain more insight in the effects of PAs and NPs in ambulance care.

Strengths and limitations
A strength of this systematic review is that the search began with a broad strategy for six online databases, following the quality standards from the Cochrane Handbook 13 and reported to conform with the PRISMA statement 28 .
One limitation of this review lies in the fact that our broad search strategy produced only four studies that described NPs or PAs working in ambulance care. Within these studies, the settings are completely different which made it impossible to perform a meta-analysis. Due to the diversity of the professionals working in ambulance care worldwide, it was difficult to identify the educational level of the professionals. Another limitation concerns the quality assessment tools for quantitative and qualitative designs a variety of these tools exists without a clear evidence base 13 .

Conclusion
This review shows that there is limited evidence on activities of NPs and PAs in ambulance care. Results show that NPs and PAs in ambulance care perform activities that can be categorized into the CanMED roles of Medical Expert, Communicator, and Collaborator. The effects of NPs and PAs are minimally reported in relation to process of care and resource use, focusing on non-conveyance rates, referral and consultation, on-scene time, or follow-up contact. There is no evidence on patient outcomes of the substitution of NPs and PAs in ambulance care. Further research is necessary to provide insight into these effects.

Data availability
Underlying data All data underlying the results are available as part of the article and no additional source data are required. The authors could have discussed this more thoroughly.

Extended data
The limitations mention the quality assessment tools used. This comes as a bit of a surprise. Also, given that only quantitative designs were found (according to table 1) I am not sure how the lack of an evidence base for the tools for assessing qualitative designs can be a limitation in this paper.

5.
Overall, I wonder if doing a systematic review limited to scientific papers should have been the design of choice for this review. Perhaps also looking at policy documents etc could have provided a richer overview of the positioning of these professionals in the field of ambulance care. This could have been mentioned in the discussion.

6.
Are the rationale for, and objectives of, the Systematic Review clearly stated? Partly

Is the statistical analysis and its interpretation appropriate?
Not applicable

Are the conclusions drawn adequately supported by the results presented in the review? Yes
Competing Interests: No competing interests were disclosed.

Reviewer Expertise: Nursing Science
I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
Author Response 23 Dec 2020

Risco van Vliet, Emergency Medical Service, RAV Brabant MWN, 's-Hertogenbosch, The Netherlands
Dear Lisette Schoonhoven Thank you for your valuable review of our manuscript and the critical questions you have asked. Before we started this review, we searched for available information that provides insight into NPs and PAs working in ambulance care. We found very little and what was written was often a personal story and not judged by a peer reviewer. We wanted to tackle this in a scientific way and expected that little had been written but were surprised by the low inclusions. Because of this, you can indeed question whether this is the correct design to answer this research question. Perhaps we would have emphasized this more strongly in