Effects of task shifting from primary care physicians to nurses: a protocol for an overview of systematic reviews

Abstract

Introduction

Task-shifting from primary care physicians (PCPs) to nurses is one option to better and more efficiently meet the needs of the population in primary care and to overcome PCP shortages. This protocol outlines an overview of systematic reviews to assess the effects of delegation or substitution by nurses of PCPs’ activities regarding clinical, patient-relevant, professional and health services-related outcomes.

Methods and analysis

We will conduct a systematic literature search for secondary literature in PubMed/MEDLINE, EMBASE, CINAHL and Cochrane databases. Systematic reviews, meta-analyses and Health Technology Assessments in German and English comprising randomised controlled trials and prospective controlled trials will be considered for inclusion. Search terms will include Medical Subject Headings combined with free text words. At least one-third of abstracts and full-text articles are reviewed by two independent reviewers. Methodological quality will be assessed using the Overview Quality Assessment Questionnaire. We will only consider reviews if they include controlled trials, if the profession that substituted or delegated tasks was a nurse, if the profession of the control was a PCP, if the assessed intervention was the same in the intervention and control group and if the Overview Quality Assessment Questionnaire score is ≥5. The corrected covered area will be calculated to describe the degree of overlap of studies in the reviews included in the study. We will report the overview according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Ethics and dissemination

The overview of secondary literature does not require the approval of an Ethics Committee and will be published in a peer-reviewed journal.

PROSPERO registration number

CRD42020183327.

Strengths and limitations of this study.

• This overview will include systematic reviews identified by a systematic literature search conducted in five databases.

• We will include only high-quality reviews with controlled trials using the Overview Quality Assessment Questionnaire for quality assessment and exclude reviews with a score <5.

• We will use a citation matrix to show the overlap in the included reviews and calculate the corrected covered area as a measure of the degree of overlap.

Background

Primary care, as the foundation of every healthcare system, is of great importance.¹ Primary care physicians (PCPs) serve as the first point of medical contact, act as a guide and coordinator in the health system, empower patients and provide long-term care for the chronically ill.² Furthermore, strong primary care is associated with better health outcomes.³ 17 out of 24 European countries, including Norway, report a shortage of medical doctors.⁴ In many OECD (Organisation for Economic Co-operation and Development) countries, there are concerns about a shortage of medical doctors, particularly general practitioners and doctors in rural areas. This is because there is often a high concentration of doctors in national capital regions (eg, Austria, the Czech Republic, Denmark, Greece, Hungary, Portugal, the Slovak Republic and the USA).⁵ Possible shortages of PCP may make it challenging for some countries to maintain high-quality primary care and provide sufficient coverage in all regions. This could be a challenge especially in regions where vacancies are particularly difficult to fill, such as rural areas.^{6 7} A simulation by WHO in 2017⁸ showed different scenarios for countries regarding the health workforce based on this simulation, certain countries may experience a surplus in nurses but a shortfall in physicians, such as Austria, Hungary, Japan, Republic of Korea, Slovakia and Slovenia. In contrast to this, other countries may experience a shortage of both professions (eg, USA). Projections for PCPs predict a shortage of between 17 800 and 48 000 physicians in the USA by 2034.⁹ However, we do not know how numbers in professions will develop actually, as the simulations underly several assumptions and are sensitive to differences in these assumptions.⁸ Moreover, many countries try to counteract these challenges with a number of measures. There is already some work available that proposes measures.^{6 10}

However, nurses are the largest workforce in most healthcare systems.^11–13 Their roles in the healthcare system include health promotion, disease prevention, care of physically ill, mentally ill and disabled people in both inpatient and outpatient settings.¹⁴

Although physicians and nurses are two distinct professions with different roots and philosophical backgrounds,¹⁵ task-shifting from PCPs to nurses is one option to better and more efficiently meet the needs of the populations in the primary care setting and to overcome PCP shortages. Nursing promotes self-care¹⁶ and the independent performance of activities that contribute to health or recovery,¹⁴ which aligns with some aspects of the WONCA (World Organization of Family Doctors) definition of general practice.² There are certain aspects of primary care nursing and general practice that are common to both professions, although they should be regarded as distinct.

The role of nurses in primary care varies across countries. A WHO report analysed 40 European countries in terms of their primary care and professional groups. In 17 out of 40 European countries, nurses work in primary care settings. However, the report does not specify whether nurses work directly in general practices or as community nurses or home care nurses.¹⁷ In Great Britain, there are general practice nurses (GPN) who work on different levels, including practitioner, senior level, advanced nurse practitioner and consultant practitioner. Typically, they hold a bachelor’s degree with a special education for general practice. Depending on the level, nurses make a significant contribution to primary care, with a wide range of competencies including the promotion of healthy lifestyles, vaccination programmes, performance of diagnostic procedures such as ECG and spirometries and therapeutic procedures such as wound care, cannulation of venous lines, prescription of medication, follow-up of patients with chronic pathologies, including health education.^18–21

In the primary care setting, there are two main professional nursing groups: practice nurses (or primary (health) care nurses) who primarily support general practitioners, and advanced nurse practitioners, who have an extended area of expertise based on in-depth education, usually at the master’s level. They sometimes also work in a general practice-substituting position.²²

Previous systematic reviews have demonstrated that substituting PCP tasks with nurses could improve outcomes.^{23 24} The assessed interventions included preventative and curative measures, urgent consultations and treatment of patients with chronic conditions. The evidence was of low to moderate quality.²³ Overviews of systematic reviews on similar topics have been published. They however focused on more occupational groups eligible for task-shifting and the methods used were more open to different study designs.²⁵ In addition, the definition of the primary care setting was broader.^{25 26}

There is a lack of a comprehensive review that evaluates preventive, curative, rehabilitative and palliative interventions in primary care, especially those provided by nurse substitutes for PCPs, using a rigorous methodology based on prospective controlled trials and reporting the results as comprehensively as possible, so that decision-makers in different healthcare systems can make decisions based on a high level of evidence. We also know little about the education that nurses receive before they start substituting. It is unclear whether the education of GPN is academic, leading to a bachelor’s, master’s or doctoral degree or whether it is specific, non-academic education and whether and to what extent the education includes practice.

Aim and review questions

The aim of the overview of systematic reviews is to assess the effects of delegation or substitution of PCPs’ activities by nurses on clinical, patient-relevant, professional and health services-related outcomes and to provide an overview of all primary care tasks that can be performed by nurses.

The review questions are:

1. Which tasks and complex interventions usually carried out by PCPs are delegated to or substituted by nurses?

2. What are the effects of delegation or substitution by nurses of PCPs’ tasks and interventions compared with routine care provided by PCPs regarding

• clinical surrogate outcomes,

• patient-relevant outcomes,

• professional outcomes and

• health services-related outcomes?

3. What is the education of nurses who carry out medical activities in the trials?

Methods and analysis

Protocol and registration

The study protocol has been registered on PROSPERO (https://www.crd.york.ac.uk/PROSPERO). The first registration date, and thus the start date, was 7 July 2020. Due to staffing constraints, we did not complete the review in 2020, but restarted the review in 2023. The expected completion date for this overview is 30 March 2024.

This protocol is reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)-Protocol 2015 statement^{27 28} (see online supplemental 1).

Patient and public involvement

No patients were involved.

Search strategy

The following databases are searched: PubMed/MEDLINE, EMBASE, Cochrane Library and Cumulative Index of Nursing and Allied Health Literature (CINAHL). Selected terms related to the topics ‘nursing’, ‘substitution’, ‘family physician’ and ‘primary care’ are applied using Medical Subject Headings and free text words. The date of the last search was 23 June 2023. Our search is limited to English and German. The full search strategy is available in the supplement (see online supplemental 2). The search is limited by date from 2020 to 2023 because we have already conducted a search in 2020 and were unable to complete the review due to staffing constraints. We are therefore carrying out an updated search. Our initial search in 2020 was limited to systematic reviews published since 2005. We chose this limitation because it seemed appropriate to us, as medicine, education and healthcare systems and their regulations are constantly evolving. This allows the authors of more recent reviews to better reflect today’s circumstances.

Inclusion criteria

Inclusion criteria are listed in table 1.

Table 1.

Inclusion criteria

Study design	Systematic reviews, meta-analyses or health technology assessments which included randomised controlled trials and prospective controlled trials
Language	English, German.
Population	Patients being treated in primary care (adults and children).
Intervention	Medical task or intervention performed by nurses (delegation or substitution).
Control	Medical task or intervention performed by primary care physicians.
Outcomes	Clinical outcomes (eg, blood pressure levels, vaccination coverage, cholesterol levels).  Patient-relevant outcomes (eg, satisfaction, hospitalisation, same-day consultation).  Professional outcomes (eg, job satisfaction, reduced workload, education).  Health services-related outcomes (eg, costs, efficiency, prescription expenditure).
Setting	Primary care (results on primary care have to be presented).

We will only consider reviews:

• If the reviews included studies with controls (randomised controlled trials or prospective controlled trials).

• If the setting and the study results were clearly assignable to the primary care setting.

• If the profession that substituted or to whom tasks were delegated tasks was a nurse or study results were reported separately and assignable to nurses.

• If the control profession was a PCP.

• If the intervention studied was the same in the intervention group (nurses) and the control group (doctors) (additional services performed by nurses compared with routine care by doctors providing not the same scope of service were not considered).

• If the publication provided all relevant information (PRISMA criteria were fulfilled; conference abstracts were not considered).

Exclusion criteria:

• Any task or intervention which is additionally performed to usual care and/or is carried out in an inpatient setting or outpatient secondary care setting.

Screening

A total of four researchers will be involved in the screening process (MP-A, NP, NPL and TS). Of these four people, three are experienced in systematic reviews. The abstracts and full-text articles will be screened independently by two randomly assigned reviewers (from the above-mentioned four researchers). In order to achieve a high level of agreement in the selection of the studies, we will hold a meeting after the first 50 screenings of each person. This will ensure a standardised approach and, if necessary, the inclusion and exclusion criteria will be specified. If we have an agreement rate >90% between the two reviewers after screening one-third of the abstracts, the remaining abstracts will be screened by one experienced scientist alone. Although double screening is recommended, single screening by an experienced researcher may also be considered appropriate.²⁹ As there is no external funding for this project, this approach is planned.

In the event of disagreement between two reviewers, a consensus will be reached by discussion. In the case of uncertainty in a single review, another experienced reviewer will be consulted. Full-text articles will only be considered if they comply with the PRISMA checklist.³⁰ Reasons for the exclusion of full-text articles will be reported. A PRISMA statement flowchart will be used to summarise the selection of studies. If an updated review has been published, we will include the most recent version. Conference abstracts will be excluded because full information according to the PRISMA statement is not available.

Quality assessment

Methodological quality will be assessed using the Overview Quality Assessment Questionnaire (OQAQ).^{31 32} We will choose the OQAQ because we started the first registration in 2020, but had to stop due to staffing constraints. At the time, we followed the IQWiG methods paper (V.5.0),³³ which listed the OQAQ, AMSTAR (Assessing the Methodological Quality of Systematic Reviews) and ROBIS (Risk of Bias in Systematic Reviews) as possible instruments. The OQAQ is a 10-item questionnaire designed to measure the ‘scientific quality’ of review articles (eg, whether the search for evidence was reasonably comprehensive or whether the criteria used to assess the validity of the included studies were reported). In total, a systematic review can receive up to seven points in this assessment tool. If it receives more than five points, it is categorised as having ‘minor or minimal flaws’.³² Therefore, only reviews scoring 5 or more points will be included in the data synthesis and the score of the included reviews will be reported. A methodological analysis of meta-analyses of analgesic interventions showed that about half of the included meta-analyses would meet this criterion (≥5 points), and the median overall score was 4.³⁴ In contrast, a paper evaluating systematic reviews in regular journals had a median score of 1, indicating extensive flaws and Cochrane systematic reviews had a median overall score of 6.³⁵

In order to achieve a more consistent rating, the OQAQ will be supplemented by a list of examples of explanations for the rating, compiled by some of the authors on the basis of their many years of experience. To determine whether there is reporting bias, we will check whether the protocol of the review has been published. We will review the PROSPERO registry.

Data extraction

Data will be extracted according to a structured extraction table. The extraction table will be piloted on two relevant studies to ensure applicability and will be updated as necessary. One reviewer will extract the data and a second reviewer will check or correct the extracted data. Disagreements will be resolved by discussing the conflict or by involving a third reviewer.

The extraction table will include the following characteristics of all included reviews: title, first author, year of publication, date of search, databases, inclusion criteria, number of included studies, OQAQ score, description of the intervention, intervention period, nurse education and outcome measures. We will pay particular attention to information on nurse education. If reviews with a broader research question are included, only the results relevant to our research questions will be extracted. If the systematic reviews do not adequately report on the interventions, the description of the intervention will be extracted from the included primary studies. These results will be published in the results section or the appendices of the final manuscript.

Management of overlapping systematic reviews

We will include all relevant reliable Cochrane and non-Cochrane systematic reviews in the overview regardless of the degree of overlap using the full inclusion technique.³⁶ As an overview of systematic reviews may include reviews that include the same primary studies, it is very important to report overlap. Failure to consider overlap may lead to bias in the results. We will use a citation matrix to show the overlap in the included reviews and will calculate the corrected covered area (CCA) as a measure of the degree of overlap according to Pieper’s suggestion: CCA 0–5 (low), 6–10 (moderate), 11–15 (high), >15 (very high).^{37 38} In addition, there are several ways of presenting overlap using tables and graphs³⁹; we will try to find an appropriate and easily understandable way of presenting overlap in our review of systematic reviews. We will not carry out a quantitative synthesis and will draw conclusions carefully, taking into account the overlap.

Data synthesis and analysis

The data synthesis will be based on a narrative synthesis. We will report the characteristics of the included reviews in a table and provide all relevant basic information about systematic reviews. We will outline the countries where the original studies were conducted and the nurses’ education. We will provide a list of specific tasks or interventions that nurses have substituted with equal or better outcomes compared with PCPs. We will also provide a list of specific tasks or interventions that nurses have substituted with worse outcomes.

The primary outcomes of our overview are patient-relevant outcomes, including symptoms, complications/adverse events, survival/mortality, pain, quality of life and patient satisfaction. As a guide to what will be defined as a patient-relevant outcome, we will use the results of the scoping review with reported and justified patient-relevant outcomes.⁴⁰

We will report all prespecified outcomes, regardless of the statistical significance of the results. We will not reanalyse outcomes data for any review or subpopulation. We will stratify the reporting of outcomes by subgroup when deemed appropriate. We will report the overview according to the guidelines in the PRISMA statement⁴¹ and a checklist developed by Li et al.⁴²

Meta-analyses will not be carried out because of the likely high heterogeneity and wide variability of the interventions studied.

Discussion

We will perform an overview of systematic reviews to assess the effects of delegation or substitution of PCPs’ activities by nurses on clinical, patient-relevant, professional and health services-related outcomes.

The strength of this overview will be that we will only include high-quality reviews with controlled trials using the OQAQ rating for quality assessment. We will focus only on nurses in the intervention group and accept only PCPs as the control group because we want to compare only these two professions. Results will be reported comprehensively, taking into account relevant influencing factors, including education.

A potential limitation of this overview will be its complexity. The topic analysed is a complex issue in several respects:⁴³

• Intervention complexity: There may be medical tasks that could be described as a single intervention, such as performing an ECG (echocardiogram), but there are also complex interventions such as disease management programmes or therapies over a longer period of time. These would have more than one or two components, and therefore a comparison between the two professions might be more complicated than for a single intervention.

• Implementation complexity: There may be different levels of education for the nurses participating in the studies. In addition, the role of nurses in the healthcare system, and thus the basic scope of practice, may vary depending on the country in which each study was conducted. This complicates the interpretation of the results. However, we will try to interpret the results of the studies by level of education. Evidence-based findings on education are important for the future planning of nursing educational programmes, which is why we are also interested in details on the education of the study nurses. We want to find out which medical activities can be carried out by nurses and to what effect, and what education is required for this (eg, academic education or clinical training, how many hours, full-time or part-time).

• Context complexity: We may detect nurses who are that are specialist nurses, for example, for dermatology, as well as GPN who could completely take over the tasks of the physicians. There will be a difference in interpretation if the nurse is responsible for only one or more indications.

• Complexity of participant responses: The acceptance of the nurses by the patients in the study may also have an impact on the results.

All these factors may make it difficult to synthesise the results. The results must therefore be presented in detail and the conclusions drawn with caution. Because of the wide range of services substituted by nurses, we expect to get heterogeneous results in terms of interventions and reported outcomes.

Implications for practice

This systematic review should be used by decision-makers (eg, in healthcare systems or primary care centres) to show what evidence is available. The translation of the results into practice in each individual country depends on several influencing factors: the number of doctors and nurses, the education and legal competencies of nurses, the willingness of physicians to give up and of nurses to take on new tasks and acceptance of the patients.

By taking on medical tasks, we expect nurses to do more than just medical work. As nurses have a different philosophical background¹⁵ and their key roles include advocacy, promotion of a safe environment, participation in patient management,¹³ they will have a fundamentally different approach to the work, for example, health promotion, attention to fragility, fall prevention and dependence prevention.

It will be necessary for the decision-makers in each healthcare system to make an individual decision on task-shifting based on their regional and country-specific healthcare situation.

Ethics and dissemination

This review does not require ethics committee approval because it uses only published studies. To the best of our knowledge, this is the first systematic review that directly compares nurses and PCPs and includes only high-quality reviews. This work will provide a list of tasks that can be substituted by nurses for decision-makers and stakeholders in decision-making, recognising their education. We will publish our findings in a peer-reviewed journal.

Ethics statements

Patient consent for publication

Not applicable.