We conducted this integrative systematic review using Joanna Briggs Institute (JBI) methodological guidance [22] and following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analysis) guidelines [23] (Multimedia Appendix 1). This type of systematic review allows quantitative and qualitative data resources to be extracted and synthesized [24]. The term integrative systematic review is often used interchangeably with mixed studies review [25]. Integrative systematic reviews have the potential to contribute to nursing theory development, informing research, practice, and policy initiatives [24].

We developed a detailed search strategy informed by previous literature [8,13,26,27] and refined by a research librarian (Y Zhang) (Multimedia Appendix 2). There are three main parts in the search strategy: (1) terms related to “palliative care,” including the clinical content of palliative care, such as symptom management, spiritual support, grief support, etc; (2) terms related to “telehealth”; and (3) terms related to “nursing.” The terms in each part are linked with “OR,” and the terms among the 3 parts are linked with “AND.” We included literature from PubMed, Embase, Web of Science, CINAHL, and Cochrane Library from January 1, 2014, to May 2024. We also manually searched reference lists of included studies.

Inclusion and exclusion criteria were formulated following the SPIDER (Sample, Phenomenon of Interest, Design, Evaluation, and Research type) framework, as detailed in Table 1 [28]. English-language; peer-reviewed; and primarily qualitative, quantitative, and mixed methods studies that focused on nurse-delivered telehealth in home-based palliative care were included. Given the rapid advancements in telehealth technologies, studies conducted too early may have limited relevance to technological contemporaneity and clinical applicability of findings. Therefore, we restricted our literature search to studies published within the past decade (from January 1, 2014, to May 2024).

The target population included adult patients (aged 18 years and older) receiving palliative care, as well as their families, caregivers, and health care providers (mainly nurses), according to the International Association for Hospice and Palliative Care [29]. The phenomenon of interest included telehealth home-based palliative care delivered by nurses.

Patients were excluded if they had an active suicidal ideation. In addition, protocol proposals were excluded because it is difficult to assess influence factors before implementation in the real world. Telehealth technologies only used to collect data were excluded because the application of telehealth technologies did not refer to intervention or implementation.

The titles and abstracts were screened in the first round, and the full texts were screened in the second round. Each round was screened by 2 researchers (CM and HZ) independently. Any discrepancies were resolved by consensus among the 2 researchers and the senior researchers (Y Zheng and Y Zeng). EndNote 21.2 (Clarivate) and the web-based program “Covidence” were used to manage references. Covidence facilitated references screening.

The methodological quality of the studies was assessed independently by 3 researchers (CM, YF, and WZ) using the Mixed Methods Appraisal Tool (MMAT; version 2018) [30]. Any discrepancies were resolved by consensus between the 2 researchers and the other senior researcher (Y Zeng). The MMAT is a critical appraisal tool designed for the quality assessment of systematic mixed studies reviews and integrative systematic reviews, which uses 5 criteria to score each study [30]. The development team of the MMAT did not advise calculating an overall score from the ratings of each criterion or excluding studies based solely on low methodological quality [30].

The entities to extract were identified by research questions of interest regardless of the type of study, which included the characteristics of articles (author, publication year, country, program name, aim, main findings, and quality assessment) and the study design of primary research (methodology, setting, sample, outcome, and measurement). We also extracted the innovations’ elements (why, what, who delivered, to whom, how, where, when, how much, and how well) according to the TIDieR (Template for Intervention Description and Replication) checklist [31]. Furthermore, 2 researchers extracted the entities and a third researcher checked the accuracy of the data (MC, FYF, and ZWC). Researchers contacted the corresponding authors to obtain information that was not available in the literature by email or ResearchGate. Data were captured across multiple Microsoft Excel spreadsheets. The roles of nurses in delivering telehealth home-based palliative care are synthesized according to extracted data.

Guided by JBI methodological guidance, we applied a convergent synthesis approach to integrate quantitative and qualitative data, which involves transforming data into a mutually compatible format [22]. We also applied qualitizing, one of the data transformation methods to convert quantitative data into themes, categories, typologies, or narratives [22]. Qualitizing involves extracting data from quantitative studies and translating or converting it into textual descriptions to allow integration with qualitative data [22]. The transformed quantitative data and qualitative data are assembled simultaneously.

First, the quantitative data from both the quantitative study and the mixed methods research were transformed into qualitative data in the form of a narrative summary [22]. Second, the qualitative data (ie, results section) from the qualitative study and the mixed methods research, along with the transformed quantitative data, were imported into NVivo 14 (Lumivero). Third, guided by the updated Consolidated Framework for Implementation Research (CFIR) 2.0, the facilitators and barriers to implementing nurse-delivered, telehealth, home-based palliative care were synthesized. The flow diagram is detailed in Figure 1. Adopting a hybrid deductive-inductive approach, we synthesized qualitative data at three phrases: (1) initially, we used the themes identified by the authors to code the primary qualitative data; (2) then, we extracted secondary themes under the corresponding constructs of CFIR 2.0; and (3) finally, we categorized these secondary themes into different levels. The results of qualitative studies included 3 elements—author-defined themes, author’s description of themes, and primary data [32].

The CFIR is one of the most commonly used determinant frameworks to assess contextual factors that influence implementation in the real world [21], which is appropriate for exploring the interesting questions of this review. CFIR 2.0 contains 48 constructs and 19 subconstructs across 5 domains (innovation domain, outer setting domain, inner setting domain, individuals domain, and implementation process domain), with the individuals domain including 2 subdomains (roles subdomain and characteristics subdomain) [21]. The characteristics subdomain is based on the capability, opportunity, motivation, behavior system or role-specific theories [21], which could document the characteristics applicable to the roles of nurses in telehealth home-based palliative care.

The electronic search was accomplished on May 14, 2024. We identified 4819 unique articles, and 1560 duplicates were removed on Covidence because they were repeated in different databases (Figures 1 and 2). We included a total of 34 articles, encompassing 29 unique primary research (Multimedia Appendix 3 [33-66]). The main items of the included studies are detailed in Table 2.

The included studies were conducted in the United States (n=11) [33-47,67,69], the Netherlands (n=3) [48-50], Canada (n=3) [51-53], Italy (n=2) [54,55], Iran (n=2) [56,57], Australia (n=1) [58], Denmark (n=1) [59,60], Switzerland (n=1) [61], Norway (n=1) [62], the United Kingdom (n=1) [63], China (n=1) [64], Lebanon (n=1) [65], and India (n=1) [66].

We included 6 qualitative studies—qualitative description (n=5) [43,51,57,62,65] and phenomenological methodology (n=1) [59,60]. A total of 17 qualitative studies were included—randomized clinical trials (n=8) [33,34,36-38,42,44,47,49,55,56,67], nonrandomized clinical trials (n=3) [39,54,55], cross-sectional studies (n=2) [50,64], and a prospective study (n=1) [66]. In addition, there were 9 studies with the design of mixed methods research—convergent parallel design (n=5) [45,53,58,61,63], explanatory sequential design (n=3) [40,41,48,52], and multiphases design (n=1) [46].

In total, 10 articles met 4-5 MMAT criteria, 23 articles met 2-3 MMAT criteria, and 1 article met 0-1 MMAT criteria. A more detailed presentation of these studies’ quality assessment is in Multimedia Appendix 4 [33-66].

Most of the included qualitative studies applied qualitative description to explore the targeted question, rather than applying a specific methodology [43,51,62,65]. The most common comments were insufficient reports on interview guide design, such as question lists [62,65]. There was a study collecting data applying both individual interviews and focus groups, without reporting the integration of qualitative data [62].

The main quality comments of included randomized controlled trials (RCTs) focused on the random allocation method, blind methods, sample size, and intervention adherence. Some of the included RCTs did not explicitly report random allocation methods [41,44] and allocation concealment measures [37,47,56]. Some studies applied a nonblinded method [49] or a single-blinded method [33,37], risking performance bias, and detection bias. Outcome assessors were blinded in some RCTs, but participants and intervention providers were not [36,67]. Some RCTs did not describe blind methods [34,42,44], limiting research transparency. Some RCTs reported a reduced sample size [33], high attrition [49], and early termination [41], affecting the accuracy of the study and reliability. Some RCTs reported low adherence [37,67] or did not report complete adherence data [34,44,47]. The most common comments on non-RCTs were no adjustment for confounders in analysis and no adherence data [39,54,55]. As for the quantitative descriptive study, the sampling method was not described [38].

Most of the mixed methods research studies did not explicitly describe integration strategies and presented quantitative data and qualitative data separately [40,48,52,53,58,61,63]. There is also a study that did not address potential discrepancies between qualitative and quantitative findings [46].

As the main delivers of telehealth home-based palliative care services delivery, nurses take various responsibilities, which are categorized by professional level and specialization level. According to the professional level, the nurses involved in these included studies were all registered nurses. On this basis, there were also advanced practice nurses (n=2) [33,34,43], nurse practitioner (n=3) [45,49,51], and clinical nurse specialist (n=1) [53]. Nurses could be categorized into 3 types based on specialization, that is, palliative care nurse (n=12) [39,43,46,49,51-54,58-62,66] or hospice nurse (n=2) [37,38,45], home care nurse (n=5) [40,41,48,53,61,66] or community nurse (n=2) [58,59], and specialist nurse (n=2) [55,62]. They undertook the responsibilities associated with the role of nurse coach (n=4) [33,34,36,44,46,67], nurse case manager (n=3) [40,41,54,55], and nurse coordinator (n=1) [62].

The contents of nurse-delivered, telehealth, home-based palliative care services involved assessment and screening (n=7) [33,34,36,39,44,46,51,53,67], palliative care consultation (n=13) [33,34,36,45,49,51,53,54,56,58-61,63,65,67], nursing coaching session (n=5) [33,34,36,42,44,46,67], regularly follow-up (n=7) [33,34,36-38,42,51,56,62,67], coordinating medical resources (n=12) [33,34,36,39,42,43,47,49,52,54,58,61,66,67], 24×7 services (n=2) [45,65], remote monitoring (n=4) [37,38,52,55,62], educational support (n=3) [39,47,48], technological support (n=5) [40,41,53,55,58,62], and home care (n=5) [40,41,48,53,61,66].

Guided by CFIR 2.0, the qualitative data and the transformed quantitative data were integrated and synthesized to identify the facilitators and barriers in 6 implementation levels and 20 constructs for implementing nurse-delivered, telehealth, home-based palliative care, as detailed in Figure 3 and Table 3.

By synthesizing quantitative and qualitative data, this integrative systematic review identified the roles of nurses in telehealth home-based palliative care and synthesized the multilevel factors to implement nurse-delivered, telehealth, home-based palliative care. Building on these findings, we have conducted a comprehensive analysis to formulate evidence-based recommendations for future implementation.

With the development of telehealth technologies, the places of palliative care services have extended to the home setting, and the roles of nurses have consequently changed [71]. This presents new challenges for future nursing competencies and practice. Due to the shortage of specialized palliative care staffing in home-based palliative care, we find that nurses’ responsibilities partially overlap with those of physicians, medical social workers, and technical personnel. Research suggests that ambiguous professional boundaries may increase the complexity of nursing practice and create role ambiguity, which may influence nurses’ own professional identity [72]. However, with telehealth-enabled empowerment, nurses could actively rebuild a stronger professional identity through fluid role boundaries [73] and facilitate communications and cooperation with multiprofessional teams [59,61]. With a firm professional identity, the nurses still can develop new roles even if the existing role collapses [74]. Similarly, a systematic review and meta-analysis found that developing the intrinsic motivation of nurses such as professional identity development could improve nurses’ job satisfaction effectively, which could reduce turnover of nurses [75].

The formation of nurses’ professional identity is a continuous process that occurs throughout nurses’ careers from nursing students’ education to continuing education [74]. It is suggested that palliative care training should be integrated into nurse education, including nursing students, primary care nurses, and specialist palliative care nurses [76]. A survey has shown that nursing students’ professional identity is related to death anxiety, and palliative care education could help them relieve death fear and develop a higher professional identity [77]. It is consistent with our findings that after receiving professional palliative care training, improvements in the capability of empathy, communication skills, and self-care techniques could help nurses reflect on and negotiate conflicts within their roles [62]. In addition to palliative care competence, we find that telehealth competence is also important for nurses to deliver telehealth home-based palliative care. One of the included studies indicated that the degree of innovation in self-efficacy of palliative care nurses significantly affects their telehealth readiness, and nurses with higher levels of innovation self-efficacy and telehealth readiness have a stronger willingness and motivation to apply telehealth technologies in home-based palliative care [64]. The telehealth competence of nurses also includes the ability to evaluate patients’ needs and willingness to use telehealth services, use telehealth devices and applications, solve technical problems, interact with patients, and improve their competence constantly [71]. A comprehensive telehealth training system should be established for palliative care specialist nurses [64,78].

Although nurses make up the largest group of health care providers in many countries and regions [79], there is a point of review that the nursing discipline has insufficiently communicated nurses as influential leaders [80]. We find that palliative care nurses are recognized and given the chance to lead the delivery of telehealth home-based palliative care in both nursing teams and multiprofessional teams [81]. The professional identity of palliative care nurses is further established in interpersonal communication and teaming among professional health care providers [74,81]. We find that nurses are deeply involved in all aspects of telehealth home-based palliative care implementation. In some countries and regions with low resources, nurses could advocate for nurses’ prescriptive authority at the policy and legal level, to improve the accessibility of medication management for patients in telehealth home-based palliative care [18]. With the transformative development of telehealth technologies, nurses could create nurse-led implementation solutions and influence actions via telehealth in multilevel settings, including nursing practice, academic research, clinical management, policy promotion, and public health in home-based palliative care [80,82]. Furthermore, nurses could lead the development of nursing education, standard of care, clinical guidance, implementation manual, quality control, and evaluation systems for the application of telehealth technologies in home-based palliative care services to improve patient safety [19,83]. As telehealth delivers, nurses also provide valuable contributions to the interdisciplinary cooperation of nursing and telehealth [84].

The core of palliative care is consistent with the orientation of the nurse-led model of care, which is to deliver patient-centered holistic care [19]. The professional identity of nurses is also established in the trust-based nurse-patient relationship [74]. Telehealth technologies transform the nurse-patient interaction [71]. Some nurses are concerned that virtual palliative care limits their ability to form personal connections with patients and their caregivers, due to lack of physical contact and incomplete assessment [51]. It also reflects an ethical criticism that telehealth home-based palliative care lacks humanity [70]. However, our review finds that from the perspectives of patients, telehealth seems to have the potential to promote the building of trust-based nurse-patient relationships [59,63]. First, nurse-delivered care could relieve patients’ symptom burden and improve their quality of life, indicating that the professional competence of nurses in palliative care could increase patient trust [15]. Second, patients who are homebound could seek timely assistance from nurses through telehealth technologies when they encounter acute symptoms, which relieves patients’ existential anxiety, thus promoting their trust in nurses [55]. Third, with the emotional and technical support of nurses, patients are empowered with self-efficacy in telehealth [43]. Furthermore, nurses also empower the patients to engage in self-care and enable their family caregivers to be equally involved in home care through health education [18]. In conclusion, with the transformation of the nurse-patient relationship, nurses need competence in interacting with patients by using telehealth technologies [71].

This study explores facilitators and barriers influencing nurse-delivered, telehealth, home-based palliative care services from an implementation science framework, without evaluating intervention effectiveness. The included studies used multidimensional outcome measures, encompassing clinical outcomes, health care resource utilization, implementation outcomes, and experiences (patients, caregivers, nurses, and other stakeholders). Among these, quality of life and mood emerged as the most frequently assessed outcomes for patients and family caregivers. Future studies could employ systematic review and meta-analysis to examine the effectiveness of innovations.

Heterogeneity in outcome measures limiting comparability. Furthermore, the lack of racial and ethnic diversity limited the generalizability of the study. There are also potential biases from including only English studies, which might affect the comprehensiveness.

This integrative systematic review synthesizes evidence on nurses’ evolving roles in telehealth home-based palliative care and identifies multilevel facilitators and barriers to nurse-delivered, home-based palliative care implementation. With the empowerment of telehealth technologies, nurses could establish a stronger professional identity and develop leadership in home-based palliative care. Systematic palliative care and telehealth education and training are critical across the nurses’ professional development, which facilitates to build nurses’ competence in home-based palliative care and develop trust-based nurse-patient relationship. Nurses are supposed to leverage influence to promote nursing practice, clinical management, and policy support in the implementation of telehealth home-based palliative care.