Background

Primary care is the foundation of a highly functioning health care system and provides comprehensive, patient-centered care that considers the needs and experiences of the individual patient, their families, and the well-being of the broader community [1]. Primary care providers are the first contact and principal point of continuing care for patients within the health care system, and coordinate other specialist care and services that patients may need [1, 2]. The delivery of primary care occurs across varied settings but is most frequently provided in a clinic and, increasingly, by interprofessional teams that may consist of family physicians, registered nurses (RNs), nurse practitioners, pharmacists, and other health professionals. In primary care settings, RNs function as generalists and provide a broad range of patient services across the lifespan, including preventative screening, health education and promotion, chronic disease prevention and management, and acute episodic care [3,4,5,6]. Specifically, family physicians and RNs represent a key collaborative relationship within these teams, contributing to strengthened primary care delivery and improvements in the comprehensiveness, efficiency, and value of care for patients [7,8,9]. Internationally, nurses are increasingly embedded in primary care settings and are recognized as the most prominent non-physician contributor to primary care teams, although the scope and speed of implementation in this area differs across countries [10, 11]. Primary care nursing in Australia is the fastest growing employment sector, with 63% of general practices employing a primary care nurse (and 82% of this group representing RNs) [12, 13]. The World Health Organization’s report [14] on the state of the world’s nursing workforce emphasizes the need to strengthen the integration of RNs into primary care, as well as the need for further research to evaluate their impact. Global workforce data are unavailable given the variability in scope of practice and role terminology, and the lack of available information across countries. A recent review of the international literature identified that titles used to refer to RNs in primary care vary across countries [15]. For instance, titles for this role in Canada are ‘family practice nurse’ and ‘primary care nurse’, whereas in Australia, the United Kingdom, and Netherlands this title is known as ‘general practice nurse’ [15]. For the purpose of this manuscript, ‘primary care RN’ will be used throughout.

Most research in this area to date has focused on describing the roles and activities of primary care RNs. A systematic review conducted by Norful et al. [5] synthesized 18 studies from eight countries related to primary care RNs and identified assessment, monitoring, and follow-up of patients with chronic diseases as fundamental roles of the primary care RN. In contrast, there have been a number of reviews conducted on the effectiveness of nurse practitioners in primary care [16,17,18]. It is imperative that primary care RNs also begin to demonstrate their contributions to patient care within this setting. Research examining RN effectiveness has primarily been conducted within the acute care and long-term care settings and focused on staffing, role enactment, and work environment. Within these settings, there is substantial evidence demonstrating the positive effects of RN staffing on improving care and reducing adverse outcomes for hospitalized patients [24, 25].

Furthermore, select countries including Australia, Canada, New Zealand, and the United Kingdom have developed national standards of practice or competencies to define the scope and depth of practice for primary care RNs [4, 19,20,21,22,23]. National competencies for primary care RNs were recently published in Canada [7]. These competencies articulate the unique scope of practice and contributions to patient care for primary care RNs across six overarching domains, namely, (1) Professionalism, (2) Clinical Practice, (3) Communication, (4) Collaboration and Partnership, (5) Quality Assurance, Evaluation and Research, and (6) Leadership.

Theoretical foundation

Determining effectiveness normally requires an examination of an intervention (e.g., primary care nursing) on a particular outcome. Incorporation of the patient perspective offers a more complete understanding of the challenges patients face within our healthcare system, especially those with long-term chronic diseases. Measuring the patient experience, which is a strong predictor of quality and value of care, should be done systematically [26]. The Organization for Economic Cooperation and Development (OECD) Patient Reported Indicator Surveys (PaRIS) Conceptual Framework was developed through a comprehensive process involving extensive international collaborations and provides a roadmap and survey tools (i.e., patient and provider questionnaires) to focus the evaluation of health care interventions on patient-reported metrics [27]. This framework provides a fuller evaluation of performance by complimenting other metrics (e.g., system/cost outcomes), while also focusing attention on the needs of the patient. The main domains of the framework include: patient reported experience measures (PREMs), patient reported outcome measures (PROMs), and health behaviours (e.g., physical activity, diet, tobacco use, alcohol use). Within primary care, the PaRIS Framework can serve as a guide for routine collection of these outcomes to facilitate quality improvement and patient-centered care [27]. A growing body of research in this area has adapted the use of this model to serve as an organizational and methodological framework. For example, multiple studies have used this framework as a method of investigating the suitability and feasibility of questionnaire and survey instruments when addressing patient perspectives [28, 29] or in the evaluation of health-related quality of life measures from the patient’s perspective [30]. A recently published systematic review that explored the opportunities and challenges of routine collection of PREMs and PROMs data for melanoma care within primary care settings found that these measures can address important care gaps and facilitate research and assessment [31]. Similarly, a study employing qualitative methods found that the use of patient-reported measures by practitioners enhanced patients’ ability to self-manage, communicate, engage, and reflect during consultations [32]. A recent environmental scan of the PROMs landscape was conducted within Canada and internationally, indicating a lack of standardized programs for routine collection and reporting of patient outcomes. Consequently, the need for enhanced PROMs information has been identified as an area of high priority [33].

Purpose

Although existing literature provides some information about the contributions of primary care RNs towards outcomes of care, a systematic review synthesizing the effectiveness of the primary care RN workforce is needed. Prior to beginning the study, the Cochrane Database of Systematic Reviews, the Joanna Briggs Institute (JBI) Library of Systematic Reviews, and the Prospective Register of Systematic Reviews (PROSPERO) were searched and no existing registered protocols or previous systematic reviews on this topic were identified. Evaluating PREMs, PROMs, and health behaviours, as well as other patient-level outcomes, is necessary to accurately demonstrate the contribution of primary care RNs, hold them accountable for their care, and generate evidence to inform decisions and policies that impact their implementation and optimization [34, 35]. Therefore, the purpose of this systematic review is to summarize evidence examining primary care RNs’ impact on patient outcomes, including physiologic changes (via biomarkers), PREMs, PROMs, and health behaviours.

Methods

Design

A systematic review of effectiveness was conducted using JBI Systematic Review methodology [36] and findings were reported in accordance with the 2009 (and where possible, the 2021) Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) framework [37, 38]. Covidence software was used to manage and organize the literature [39] and enable a team approach for study and data review. The protocol for this systematic review is registered on PROSPERO (registration ID CRD42018090767). This paper presents a summary of findings from studies that report on patient outcomes, including biomarkers, PREMs, PROMs, and health behaviours. A full description of the methods and findings from studies that measured care delivery and system outcomes are reported in the companion paper “Effectiveness of Registered Nurses on System Outcomes in Primary Care: A Systematic Review” [40].

Search strategy

The search strategy aimed to include both published and unpublished literature. Following a limited search in CINAHL and Medline that identified optimal search terms, two members of the research team performed a comprehensive search of relevant electronic databases (see Supplementary File 1). Grey literature was identified using ProQuest Dissertations and Theses, MedNar, Google Scholar, the websites of relevant nursing organizations (e.g., International Council of Nurses, Community Health Nurses of Canada), and reference lists of included articles. There were no location or publication date restrictions on search criteria. Studies published in any year up to and including the date of article retrieval (January, 2022) were considered. Ongoing searches for grey literature included studies with publication dates up to January, 2022.

Inclusion criteria

Studies considered for inclusion reported on any quantitative study published in English with outcomes that directly measured, or were related to, an intervention attributable to a primary care RN. Only studies focused on RNs or equivalent (e.g., practice nurse, general nurse) [15] were included; if the RN designation was unclear or could not be determined based on the region of publication, the study was excluded. Studies that involved primary care RNs who underwent considerable advanced/focused training or those that exclusively examined structural variables were excluded. Full details regarding inclusion criteria are published in the companion paper [40].

Screening

Reviewers included two study authors (DR and JL) and two trained research assistants (AR and OP). All identified titles and abstracts were screened independently by two reviewers for potential study eligibility. Two reviewers independently screened full-text articles for relevance, applying pre-established eligibility criteria. Any disagreements were resolved through discussion, or by a third reviewer.

Risk of bias

The risk of bias and quality of each study was assessed using the Integrated Quality Criteria for Review of Multiple Study Designs (ICROMS) tool (see scoring matrix located in Supplementary File 2) [41]. All full-text articles that met eligibility criteria were appraised for quality by two independent reviewers. All studies that met inclusion/exclusion criteria also met the minimum ICROMS score to be included in the review.

Data extraction and synthesis

All eligible full-text studies underwent data extraction using a tool pre-designed and tested by the research team and based on the Cochrane Public Health Group Data Extraction Template [42]. Data extracted from the articles included: country and year of publication, study aim and design, description of primary care setting, patient sample sizes and demographics, details of study intervention and primary care RN involvement/role, outcome measures used to evaluate these interventions, and study results. To address the broad range of terms and descriptors used across included studies (e.g., traditional care, standard care, basic support, care delivered by anyone other than a primary care RN) and to provide clarity in the presentation of our results, we refer to all control groups as “usual care” or the “comparator group”. Outcomes were grouped in accordance with the OECD PaRIS Conceptual Framework Classification [27] into one of three categories defined by this model (i.e., PREMS, PROMs, health behaviours), and an additional category added by the research team (i.e., biomarkers) (see Table 1). Biomarkers consist of outcomes related to changes in patient health status as measured by clinical assessment (e.g., hemoglobin A1c [HbA1c] values, blood pressure, body weight). PREMs are defined as patient experience indicators related to health care access, autonomy in care, and overall satisfaction with care received, and are often assessed through self-report questionnaires or population-based surveys [27]. These outcomes can be summarized further based on patient experiences surrounding access (e.g., first point of contact), comprehensiveness of care, self-management support, trust, and overall perceived quality of care. PROMs are described as outcomes relating to a patient’s self-reported physical, mental, and social health status and can be categorized as either generic or condition-specific and applied to a broad patient population [27]. Outcomes identified on this level can be further categorized into functional status (e.g., disability, physical, mental, social function), symptoms, and health-related quality of life. The remaining outcomes were categorized according to the health behaviors classification, which includes lifestyle behaviors and actions that can contribute to a patient’s overall health status (e.g., physical activity, smoking status, dietary intake) [27]. Due to the diversity of included designs, interventions, and outcomes across studies, narrative synthesis was used to present study findings.

Table 1 Classification of patient outcomes measured in each study based on the OECD PaRIS Conceptual Framework [27] 
Full size table

Results

Figure 1 presents a PRISMA diagram outlining the results of the literature search.

Fig. 1 
figure 1

PRISMA Diagram of Literature Search. *This paper reports on studies that measured patient outcomes. Findings from studies that measured care delivery and system outcomes are reported in the companion paper "Effectiveness of registered nurses on system outcomes in primary care: a systematic review" [40]

Full size image

Study characteristics

Of the 29 articles included in the final review, 23 reported on patient outcomes (included in the present analysis) [40]. Table 2 presents a detailed summary of the study characteristics for each of these articles. Studies were published between the years 1996–2021 and conducted in the United Kingdom (n = 9), United States (n = 6), Australia (n = 5), and New Zealand (n = 3). Study designs included randomized controlled trials (n = 9), observational (n = 8) (e.g., survey, secondary data analysis), cohort (n = 1), non-controlled (n = 2) and controlled before-after (n = 1), and two studies with mixed-methods designs that combined a non-controlled before-after with a non-randomized controlled trial (n = 1) or with an observational design (n = 1). Sample sizes ranged from 81–2850 patients. Quality scores, as assessed by the ICROMS tool, varied between studies. Four studies were scored at the minimum threshold for their study design [46, 56, 61, 65], six studies scored 1–2 points above threshold [44, 45, 48, 49, 53, 57], and thirteen studies exceeded the minimum cut-off score by 3 or more points [43, 47, 50,51,52, 54, 55, 58,59,60, 62,63,64].

Table 2 Literature review table of study characteristics (n = 23)
Full size table

Overview of RN interventions

The nature of interventions examined in this review differed across studies. The most common interventions were related to chronic disease prevention and management, specifically, case management or targeted chronic disease management care (e.g., diabetes, obesity, hypertension, hypocholesteremia) (n = 7) [43,44,45,46,47,48,49] and primary and secondary preventative care for patients at risk of chronic disease (e.g. prediabetes, coronary heart disease) (n = 3) [50,51,52]. Other studies examined primary care RN-delivered smoking cessation support (n = 4) [53,54,55,56], back pain education and management [57], pelvic floor muscle training [58], consultations aimed at increasing patient physical activity levels [59, 60], and a telecare program for patients with diagnosed depression [61]. Four studies examined the impact of RN care in general (at an organizational-level); three of which focused on consultations with patients in general practice [62,63,64] and another which examined the impacts of a nurse-operated telephone consultation/triage service [65].

In thirteen studies, primary care RNs carried out the intervention independently without the support of other staff/providers [45, 46, 49, 51, 53,54,55, 57,58,59,60, 62, 63], and in 10 studies, they carried out the intervention interdependently, in collaboration with health care providers (e.g., physicians, clinical pharmacy specialists [CPS], dieticians) or members of the research team (e.g., trial nurse facilitator) [43, 44, 47, 48, 50, 52, 56, 61, 64, 65]. Three of these 10 studies involved evaluating RNs at the general practice-level and therefore are assumed to be evaluating an interdependent role involving support of other health care providers [47, 61, 64]. The presence and type of comparator group also differed across study designs. Specifically, five of the included studies compared a nurse-led intervention to the same intervention led by other health care providers [46, 52, 54, 55, 58]. Other studies compared nurse-led interventions with that of ‘usual care’ not associated with nurse involvement (n = 4) [43, 50, 57, 60], or with ‘usual care’ that was associated with reduced or alternative levels of nurse involvement (n = 5) [44, 49, 53, 56, 59]. The remaining studies examined the effectiveness of a primary care RN-delivered intervention on specific outcomes of care using an observational or before-after design (n = 5) [48, 51, 61, 62, 65], or did not contain a specific intervention, but rather, examined the impact of varying roles and practice characteristics of the primary care RN in general practice (n = 4) [45, 47, 63, 64].

Overview of outcomes

A total of 46 patient outcomes were identified across included studies (Table 3). Physiologic disease control outcomes, which were measured via biomarkers, included quality of care for diabetes (e.g., HbA1c, fasting blood glucose) [43, 44, 50, 51], obesity (e.g., body mass index [BMI], waist circumference) [44, 47, 50, 51, 59, 60], pelvic floor strength and endurance [58], hypercholesterolemia (e.g., total cholesterol) [49], and hypertension (e.g., blood pressure) [43, 44, 48, 50, 51]. Patient experience outcomes identified under the PREMs category included patient satisfaction with access to care (RN versus physician as first point of contact) [64], quality of self-management support (e.g., smoking cessation counseling, chronic disease services) [56, 62], comfort/trust with primary care RN roles [45], and overall satisfaction or perceived quality of care with provider consultations, treatment, or advice/support received [45, 51, 55, 57, 63, 65]. Patient reported outcomes identified within the PROMs category consisted of physical and social functional status [43, 57], level of disability (e.g., activity levels, bed rest, work loss) [57, 61], changes in self-reported anxiety, depression, or pain [59,60,61], adverse health events (e.g., falls, fractures, severe hypoglycemia) [43, 59, 60], and health-related qualify of life (e.g., physical activity, social activity) [43, 46, 51, 52, 60]. Lastly, outcomes grouped under the health behaviors classification included reduction and/or cessation of tobacco use [51, 53,54,55,56], changes to physical activity (e.g., level of aerobic exercise, daily step count) [51, 57, 59, 60], and improvements in dietary intake [49].

Table 3 Literature Review Table – Description of Patient Outcomes and Study Results
Full size table

Physiologic disease control via biomarkers

Ten studies measured clinical patient outcomes when comparing interventions that involved primary care RNs to that of usual care or an intervention delivered by a comparator group. Clinical biomarkers included those for diabetes (HbA1c, fasting blood glucose), obesity (BMI, total fat mass), hypertension (blood pressure), and cardiovascular risk (total cholesterol). Of the ten studies, four examined diabetic control. After one year, Aubert et al [43]. reported significant differences in HbA1c values; patients in the primary care RN case management group had a larger mean reduction (-1.7 percentage points) over 12 months in comparison to usual care (-0.6 percentage points) (difference -1.1, 95% CI: -1.62 to 0.58; p < 0.001). Additionally, patients in the intervention group had a greater decrease in fasting blood glucose than the usual care group (-48.3 mg/dL versus -14.5 mg/dL; difference -33.8, 95% CI: -56.12 to 11.48; p = 0.003). Bellary et al [44]. found a small but non-significant reduction in HbA1c among their patient sample after two years. One additional study that conducted a retrospective data analysis of clinical outcome data from patients attending an independently RN-led primary care clinic, did not detect significant changes in HbA1c between initial intake at baseline and follow-up visits at various intervals (reported as 3 months to “several years” depending on the individual) [51].

Seven studies examined obesity-related outcomes such as BMI, weight, and total fat mass. In their adjusted regression models, Karnon et al [47]. reported that high level involvement of primary care RNs in the provision of obesity-related clinical activities (in comparison to low level involvement) yielded significantly larger mean reductions in BMI (mean difference -1.10, 95% CI: -0.45 to -1.76; p = 0.001) after one year, however, there were no significant improvements in terms of the proportion of patients losing weight (mean difference 0.09, 95% CI: -0.07 to 0.25; p = 0.259). Coppell et al [50]. found a significant weight reduction (-1.3 kg) in the primary care RN-led prediabetes intervention arm compared to usual care (gained 0.8 kg) (2.2 kg difference; p < 0.001). Mean BMI and waist circumference also decreased in the intervention arm compared to an increase in the control group, however, these differences were not significant. Likewise, a third study reported that fat mass was slightly reduced at 12 months, but differences between the intervention and control groups were equivalent when the primary care RN group was compared to both postal intervention (p = 0.54) and usual care (p = 0.30) [60]. There were no significant reductions in BMI or waist circumference in the remaining four studies [43, 44, 51, 59].

Five studies investigated the impact of enhanced nurse involvement in primary care delivery on blood pressure. Bellary et al [44]. reported significant differences between groups in diastolic blood pressure (-1.91, 95% CI: -2.88 to -0.94 mm Hg; p = 0.0001) and mean arterial pressure (1.36, 95% CI: -2.49 to -0.23 mm Hg; p = 0.018), favoring the intervention (additional time spent with a primary care RN). In a second study conducted by O’Neill et al. in which the RN independently assessed blood pressure and collaborated with either a CPS or physician in hypertension case management, [48] there was a greater decrease in systolic blood pressure in patients who received care from CPS and primary care RN teams (-14 mm Hg) compared to patients receiving care from physician-directed primary care RNs (-10 mm Hg) (p = 0.04), however, there were no significant changes in diastolic blood pressure between groups. The remaining three studies found no significant changes in blood pressure when comparing a primary care RN-led intervention to that of usual care [43, 50] or from initial baseline to follow-up [51].

Total cholesterol was measured in four studies. Pine et al  [49]. reported that the mean total cholesterol level decreased by 0.29 mmol/L (11.2 mg/dL) (4.3%) from the initial physician visit to the first primary care RN visit. Following five counseling sessions by a primary care RN, the mean total cholesterol levels of all patients decreased (-0.14 mmol/L; p = 0.4). However, during the follow-up comparison study, there were no significant differences in total cholesterol improvement between the nurse-counseling intervention group and the comparison patients, and total cholesterol levels in both groups improved significantly (p = 0.002). The remaining three studies reported equivalent results in regards to total cholesterol reduction [43, 44, 50].

Patient experience outcomes via PREMS

Nine articles reported on patient experience outcomes via PREMs: overall perceived quality of care [45, 51, 55, 63, 65], self-management support [56, 62], access (first point of primary care contact) [64], comprehensiveness [57], and trust [45].

In regard to overall perceived quality of care, Halcomb and colleagues [45] found that Australian patients were very satisfied and comfortable with chronic disease care delivered by a primary care RN. This was particularly true for patients with diabetes who reported being almost three times more comfortable (38% versus 14%, p = 0.016) with their encounter than patients who consulted for other chronic health conditions. A similar study in New Zealand also revealed high satisfaction with primary care RN-delivered services overall, with increased satisfaction associated with an increased number of visits (i.e., those who had more than four previous visits to the primary care RN) after controlling for demographic factors [63]. Longer consultation time with a primary care RN resulted in higher patient satisfaction (OR = 2.50, 95% CI: 1.43 to 4.35; p < 0.01) and patient enablement (OR = 2.55, 95% CI: 1.45 to 4.50; p < 0.01) than shorter consultation time [62]. Moreover, patients who attended practices where primary care RNs worked with broad scopes of practice and high levels of autonomy were more satisfied (OR = 1.76, 95% CI: 1.09 to 2.82; p = 0.04) and more enabled (OR = 2.56, 95% CI: 1.40 to 4.68; p < 0.01) than patients who attended practices where nurses worked with more limited scopes of practice and lower levels of autonomy [62].

Patients also reported improved health, better understanding of disease diagnosis, medication, and treatment plan, and more motivation for self-management as a result of primary care RN-led lifestyle clinics focused on diabetes, smoking cessation, women’s health, cardiovascular risk, respiratory/asthma, and diet/nutrition [51]. Furthermore, patients reported positive experiences with primary care RN-led telephone consultations for acute illness [65], back pain education [57], and smoking cessation support [55, 56]. For instance, Cherkin et al. [57] reported higher satisfaction (p < 0.01) and higher perceived knowledge (p < 0.001) for patients who received a primary care RN-led educational intervention for back pain than those in the usual care group. Nearly all patients (98%, n = 385) in an Australian study [56] that examined smoking cessation behavioral support from a primary care RN rated the support provided as helpful (19%) or very helpful (79%) and indicated that they may have been more successful with smoking cessation if they had been able to have more sessions with the RN. With respect to access to care, a study by Caldow et al [64]. found that patients expressed satisfaction and preference with primary care RN versus physician consultations for minor illness as first point of contact if this resulted in a reduced waiting time, suggesting that patients would be accepting of an expanded nursing role in primary care.

Patient reported outcomes via PROMs

Patient reported outcome measures via PROMs were examined across eight studies and included health-related quality of life [43, 46, 51, 52, 60], symptoms [59,60,61], self-efficacy [60], and functional status [57].

Health-related quality of life, as measured through patient self-report, was assessed in five studies [43, 46, 51, 52, 60]. In a 12-month randomized controlled trial conducted by Aubert et al., [43] a primary care RN-led case management model of adult diabetes care was compared with that of usual care in a primary care setting. Health-related quality of life was assessed by a validated questionnaire developed by the Centers for Disease Control and Prevention for the Behavioral Risk Factor Surveillance System to assess patient perception of health status across four domains. The results demonstrated an improved perception of health status in both groups, with patients in the intervention group more than twice as likely to report improvement in health status score (mean change = 0.47) as those in the usual care group (mean change = 0.20) ( difference= 0.27; 95% CI: -0.03 to 0.57; p = 0.02). In contrast, the other four studies [46, 51, 52, 60] examining health-related quality of life did not report significant differences in regards to these outcomes, including two cluster randomized controlled trials [46, 52, 60]. One study assessed the effectiveness of three different methods of secondary prevention care of coronary heart disease (recall to a primary care RN; recall to a physician; audit and feedback) [52], while the other compared the efficacy of a primary care RN-supported physical activity intervention to that of usual care [60]. Both studies reported equivalent scores between groups on all dimensions of patient self-reported quality of life measurements. Lastly, an observational study by Marshall et al [51]. used patient satisfaction surveys to assess perceptions of Nurse-Led Healthy Lifestyle Clinics (NLHLC) in New Zealand. Using scores from the Dartmouth Primary Care Cooperative (COOP) Information charts [66], it was noted that there were no statistically significant differences in the COOP dimensions related to self-perceived quality of life from first clinic visit to last clinic visit. However, significant improvements were noted in relation to the COOP variables related to patient-perceived social activity (mean difference = -0.20; p = 0.049), change in health (mean difference = -0.42; p = 0.001), and overall health (mean difference = -0.21; p = 0.025).

Patient self-reported symptoms were measured in three studies [59,60,61] and included outcomes related to both mental and physical health (anxiety, depression, pain), as well as the occurrence of adverse health events (injuries, fractures, cardiovascular events, deaths, and deterioration of any pre-existing health problems). In an observational study of a nurse telecare intervention for adults with depression in the United States, Pearson et al [61]. found a significant improvement in mean scores on the SF-12 Mental Functioning Scale between baseline (mean = 29.9) and 6-months post-intervention (mean = 48.2) (p < 0.0001). During the same time interval, significant differences were noted on the Hamilton Depression Rating scale (14.6 to 6.5; p < 0.001), as well as the mean scores on the Work Limitations Questionnaire (70.4 to 87.2; p < 0.001), which both represent an improvement in functioning. Paired t-test results for the difference in mean scores on all three instruments were statistically significant (p = 0.0001) and the majority of patients (63%) experienced at least a 50% reduction in the Hamilton Depression Rating score at 6-months. The remaining studies to examine self-reported symptoms as an outcome were two randomized controlled studies that measured the effects of a primary care RN-delivered intervention on patient physical activity [59, 60]. Both studies assessed changes to patient self-reported levels of depression, anxiety, and pain and incidents of adverse health events. The results of both studies reported no statistically significant between-group differences in mean scores of either symptom at 3- or 12-months post-intervention. Additionally, while total number of adverse events did not differ between groups for either study, Harris et al [60]. found a significant reduction in cardiovascular events among the intervention group over the 12 month period (p = 0.04).

Patient self-efficacy was examined in a three-arm cluster randomized controlled study conducted by Harris et al [60]., in which patient-reported levels of exercise self-efficacy were examined at 3- and 12-months following a physical activity intervention. Exercise self-efficacy in this study was characterized by a patient’s willingness to set goals, create action plans, engage in self-monitoring, and seek out social support, and are directly related to long-term physical activity adherence. Findings indicated that exercise self-efficacy was significantly increased in both intervention groups at 3-months for postal group (pedometer delivered by mail) versus control (Effect Size [ES] = 1.1, 95% CI: 0.2 to 2.0; p = 0.01), primary care RN group versus control (ES = 2.3, 95% CI: 1.4 to 3.2; p < 0.001), and primary care RN group versus postal group (ES = 1.2, 95% CI: 0.3 to 2.1; p = 0.01). For primary care RN group versus control group, the difference remained significant at the 12-month follow-up (ES = 1.2, 95% CI: 0.3 to 2.2; p = 0.01), but not for the postal group versus control (p = 0.2) or the primary care RN group versus postal group (p = 0.22).

The sole study to evaluate functional status was a randomized controlled trial conducted by Cherkin et al [57]. comparing usual care, usual care plus an educational booklet, and usual care plus an educational session with a primary care RN and an educational booklet to improve outcomes of low back pain in primary care. None of the interventions had a statistically significant effect on functional status, including days of limited activity, bed rest, or work loss resulting from back pain one week after the intervention or at any subsequent follow-up.

Health behaviors

Of all the studies included (n = 23), nine considered health behavior outcomes. Among the studies examining the impact of a primary care RN-led intervention on health behaviors, it was found that tobacco use was the most documented health behavior (n = 5) [51, 53,54,55,56]. Tobacco use was examined by looking at both abstinence from smoking, as well as daily reductions or changes to smoking behaviour, and this was measured at multiple follow-up periods throughout the duration of the intervention. All five studies demonstrated positive changes in smoking-related health behaviors following either an independent [51, 53,54,55] or interdependent [56] primary care RN intervention. For example, Byers et al [54]. compared a primary care RN-led intervention with a physician-led intervention to support smoking cessation. The results show that support provided by the primary care RN was equivalent to that provided by the comparator group (29.1% versus 18.2% quit rate, respectively; p = 0.077) in supporting the patient to quit smoking. Marshall et al [51]. looked at primary care RN-led healthy habits lifestyle clinics for patients with or at risk of chronic disease within targeted populations with known health inequalities. Following the intervention, 94% of patients reported having a better understanding of their diagnosis, medication and treatement plan, and an increase in motivation to self-manage their health needs. Other studies have examined the impact of nursing interventions on patient-reported levels of physical activity [57, 59, 60]. Only one study considered adherence to healthy eating as a health outcome following a RN-led intervention in primary care [49]. In this study, Pine and colleagues evaluated the effect of a nursing intervention to support cholesterol lowering for patients diagnosed with hypercholesterolemia. To do this, primary care RNs provided a total of five counseling visits focused on nutritional education and physical activity (1-month after referral, and at 3-, 5-, 7-, and 12-months) to 82 patients with total cholesterol higher than 6.21 mmol/l. Intervention patients were already following a diet consistent with the program at baseline, however, the mean score for Section 1 of the Eating Pattern Assessment Tool (questions related to foods with serum cholesterol-raising potential) improved from 23.4 at the first visit to 20.4 at the final visit (p < 0.001).

Discussion

This systematic review presents a comprehensive synthesis of literature examining the impact of primary care RNs on patient outcomes. The findings suggest that outcomes resulting from care provided by primary care RNs are comparable and complementary to care provided by other primary care providers, specifically with respect to chronic disease prevention and management, smoking cessation, and wellness counseling. This review supports that primary care RNs deliver appropriate and high-quality patient care. There was a high level of patient satisfaction reported regarding experiences with RN-led care. Patients appear to be comfortable with RNs providing primary care services and taking on expanded roles in primary care. This is consistent with findings from other studies that have examined patient satisfaction and comfort with RN roles in primary care practices across multiple countries [67,68,69,70]. Our findings are aligned with existing evidence that has linked patient experiences of care to the level of autonomy and scope of practice of the RN in the clinical setting. A recently published study from Canada evaluated patient experiences in primary care organizations and determined that patient-reported experience was significantly enhanced in clinics in which RNs systematically followed patients and used their scope of practice to their full potential [71].

Patient experience is a strong indicator of patient-perceived quality of care and fundamental to achieving desired patient outcomes for a range of physical and mental health domains [72,73,74,75,76]. The recently developed PaRIS Framework served as a valuable tool for organizing patient-reported outcomes, however, it did not capture all patient outcomes identified within the studies included in this review. For example, clinical biomarkers, such as HbA1c and fasting blood glucose (used as a measure of diabetes care quality), were not considered in the PaRIS Framework and therefore added by the study authors during the analysis phase as an additional patient outcome category. Similarly, studies evaluating RN interventions in primary care did not consider many components of the PaRIS Framework, such as delivery system design (e.g., clinic remuneration, remote consultations), individual and sociodemographic factors (e.g., patient demographic characteristics), and health and health care capabilities, and only measured select components from the patient-reported experiences of care, health behaviours, and patient-reported outcome domains. For instance, many articles did not provide information regarding RN characteristics, such as level of education, years of experience, or specific roles/tasks that they performed in-clinic prior to the intervention that might have impacted outcomes observed. In addition, conceptual definitions of outcomes within included studies may vary and not align with meanings as defined within the PaRIS Framework. A taxonomy for RN outcomes may be useful and should be considered in future revisions and applications of the OECD PaRIS Framework.

The included studies evaluated a variety of primary care RN interventions but did not capture all roles that encompass their broad scope of practice. Commonly offered services by primary care RNs that have yet to be comprehensively evaluated include prenatal and well-baby care, therapeutic interventions (e.g., wound care, treatment of infections), preventative care (e.g., immunizations, health promotion and education) and care coordination (e.g., nursing surveillance, system navigation). Although self-management supports (e.g., smoking cessation, physical activity, diabetes, nutrition, pain management, healthy lifestyle promotion, chronic disease prevention) were examined in a few studies [43, 44, 49,50,51, 53,54,55,56,57, 59, 60], further evaluation of self-management and behaviour support interventions (included within the PaRIS Framework) offered by RNs in primary care is needed. Moreover, recently developed competencies identify that, in addition to clinical practice activities centered around patient care provision, primary care RNs engage in a wide range of non-clinical roles, such as leadership, research, and interprofessional collaboration. These non-clinical domains of practice for primary care RNs require further understanding and evaluation.

There is a general apprehension among some medical practitioners that if RNs assume more responsibilities or enhanced roles within primary care settings, high-quality care and patient safety will be compromised [77, 78]. Findings from this study show that patient care is equivalent to that of “usual care” and in many cases, produced better patient outcomes when the intervention was provided by a primary care RN. This aligns with literature in the acute care and long-term care settings [24, 25]. The findings from our review call into question concerns that RN-provided care increases risk or reduces quality of care and equally, lends support towards the efficacy of primary care RN care provision on improvements to patient outcomes. Additionally, primary care RNs are required to practice within their legislated and regulated scope of practice, regardless of the practice setting or types of clinical roles performed [79, 80].

Generally, there are methodological challenges associated with examining the contributions of a specific health care provider within the context of a team [81,82,83]. As the focus of practice and research moves towards interdisciplinary teams, it is increasingly difficult to isolate and evaluate the impact of primary care RN interventions. In addition, the roles of primary care RNs and team compositions vary across practice settings. Although this review exclusively examined studies in which an intervention was delivered by a RN, many studies were excluded because of unclear terminology surrounding the nursing designation/role (i.e., unable to discern whether nurse in study was a RN or equivalent). Furthermore, the review included only nine randomized controlled trials, which provide the strongest level of evidence, that specifically compared RN-led interventions to care delivered by other health care professionals and/or usual care [43, 52, 53, 57,58,59,60, 84, 85]. Comparator groups in studies varied considerably, impacting the ability to make comparisons across studies and limiting the generalizability of findings from this study. Despite these challenges, this review provides preliminary evidence on patient outcomes used to evaluate a variety of different RN interventions in a multidimensional health care environment. The findings from this study, coupled with an existing framework (e.g., OECD PaRIS Framework) serve as a tool to map roles and activities to outcomes and guide future evaluation of primary care RNs. Overall, as the presence of RNs in primary care increases globally, further evaluation research implementing control/comparison groups into study design and controlling for confounding factors (e.g., nurse characteristics) is needed to strengthen the evidence related to the effectiveness of RNs in primary care.

Strengths and limitations

This review provides preliminary evidence regarding the effectiveness of RNs on patient outcomes in primary care. Traditional means of measuring the effectiveness of care provision in the healthcare sector have focused mainly on the use of clinical data. A strength of this systematic review is its patient-oriented approach that assesses health outcomes from the patient perspective [27, 86]. Additional strengths of this systematic review include the application of a comprehensive search strategy and use of the PRISMA checklist in the screening process. However, despite utilizing a comprehensive search strategy, it is possible that not all relevant studies were retrieved and included in this review. Furthermore, although we conducted an appraisal using an established quality assessment tool (i.e., ICROMS), this tool presented certain challenges. For instance, although a strength of this tool is that it offered criteria to assist with the process of assigning quality scores, there is a degree of subjectivity involved in the appraisal process. In addition, the minimum cut-off scores varied across study designs and therefore, made comparisons of the quality between different types of studies difficult. Similarly, the score itself is difficult to interpret without an understanding of the tool and design matrix (limitations of articles are summarized in Supplementary Table 3). The ICROMS tool was also not designed to specifically appraise mixed methods or observational designs. The lack of consistent and available data regarding terminology used to describe RNs, or equivalent nursing titles, across countries limited the ability to include studies published in certain regions. Only studies published in the English language were included, which may limit generalizability to certain countries. High-quality research employing robust study designs (e.g., randomized controlled trials) need to be conducted to further understand the impact of RNs on patient outcomes in primary care.

Conclusions

Primary care RNs can provide patient care that is comparable and complementary to that of other primary care providers, specifically with respect to patient satisfaction, enablement, self-reported quality of life, self-efficacy, and improvements in health behaviours. This review provides preliminary evidence regarding the effectiveness of RNs on patient outcomes in primary care. Findings are applicable to researchers and other stakeholders engaged in primary care reform and can inform further integration and optimization of this role, as well as contribute to future research.