The analysis adhered to the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) checklist (Checklist 1) [51].

An information specialist from the Amsterdam Medical Center library (JGD) was consulted to help define key concepts and establish the evaluation framework, which included selecting the appropriate databases, relevant keywords, and MeSH terms. MEDLINE (Ovid) and Scopus were searched from inception on April 29, 2024, to include studies up to the search date.

The search strategy could be conceptually summarized as follows: ([GRCs] AND (([neighborhoods] AND ([chronic care model] OR ([HISs] AND [health behaviors]))) OR ([community health facilities] AND [health behaviors]))), involving the following four elements related to our research questions: (1) HISs, (2) geo-referenced contexts (GRCs), (3) neighborhoods, and (4) health behaviors. Detailed information regarding the search terms in Scopus is shown in Textbox 1. The inclusion and exclusion criteria are presented in Textbox 2. For search terms in MEDLINE, refer to Table S2 in Multimedia Appendix 1.

GRCs include (1) built environments from the 5D dimensions (density, diversity, design, distance, and destination accessibility); (2) natural environments; and (3) social environments, which encompass neighborhood socioeconomic status and neighborhood social relationships. Neighborhoods include (1) static administrative areas, (2) various delineations surrounding each participant’s address, (3) perceived neighborhoods, and (4) geographic areas encompassing all locations visited by an individual in their daily activities. For the definitions and examples of NGRCs and neighborhoods, see Table S1 in Multimedia Appendix 1 [1,9,26,52-76]. We used 3 different search strategies because some concepts can be merged and replaced by related ones. For example, HISs and health behaviors are 2 critical elements of the chronic care model, which is a valuable framework for patient empowerment, supporting self-management and enhancing health outcomes [53-56]. Consequently, substituting HISs and health behaviors with search terms pertinent to the chronic care model while connecting them with GRCs and neighborhoods constituted the second part of the search strategy. Additionally, community health facility workers may use HISs to address neighborhood-related issues. Therefore, replacing HISs and neighborhoods with search terms relevant to community health facilities (eg, community health centers and community data) while linking these terms with GRCs and health behaviors formed the third part of our search strategy.

All original peer-reviewed scientific studies in English indexed up to April 29, 2024, were included, while non–peer-reviewed studies and gray literature were excluded. Textbox 2 outlines the inclusion and exclusion criteria.

First, we screened titles and abstracts and excluded studies according to the inclusion and exclusion criteria. When uncertainty arose, studies were retained for the full-text screening stage. Second, XY and MD independently conducted a full-text review and extracted the relevant NGRCs, HISs, and behavioral interventions from the potentially included papers. Any disagreements were first addressed by referring to a predefined inclusion/exclusion checklist. If consensus was not reached using the checklist, the reviewers discussed each case, and, if necessary, a third reviewer (RC) made the final decision. Although interrater reliability was not calculated, this structured approach ensured consistent and transparent decision-making.

We extracted data from the included studies using a 7-item table. These seven items included: (1) study attributes; (2) population attributes, diseases, and health setting characteristics; (3) HISs; (4) NGRCs; (5) behavioral interventions; (6) results; and (7) conclusions and future studies. Table S3 in Multimedia Appendix 1 provides further details.

Figure 1 presents the PRISMA flowchart, which includes the identification, screening, and selection process. A total of 24 peer-reviewed studies were finally included.

We reviewed the descriptive characteristics of all 24 included papers. Basic characteristics of the studies are presented in Table S4 in Multimedia Appendix 1, including information on (1) the type, number, gender, age distributions, and language of participants, and (2) the study type and year of publication. The characteristics with appropriate properties for summarization include (1) the type, NCD, and health and computer literacy of participants, and (2) health settings. All studies were conducted in the United States. Details are provided in Table S5 in Multimedia Appendix 1.

We divided the studies by participant type as follows: 2 studies [29,77] focused solely on providers, 4 studies concentrated on caregivers [78-81], 13 studies examined patients only [10,82-93], and 5 studies involved both health providers and patients [93-97]. Among patients, most studies focused on adults.

In this scoping review, we reviewed the current state of NGRC-focused behavioral interventions using HISs in order to propose recommendations for future studies. We included 24 studies, and nearly half of the studies (11/24, 46%) were conducted in 2023 or later (Table S4 in Multimedia Appendix 1). Due to unstructured reporting in many of the included studies, reliably extracting data was difficult. We organized our findings and recommendations according to the 4 research questions.

The predominant reliance on provider-centric EHRs/EMRs and the limited and inconsistently described use of HISs across the entire NGRC-focused behavioral intervention cycle significantly constrain the integration of NGRCs into behavioral interventions. Apart from the unspecified type of HIS in 1 study [10], most included studies relied on provider-centric EHRs/EMRs (20/24, 83%), with limited use of more patient-empowered PHRs. Only 3 studies used tethered PHRs, and none utilized interconnected PHRs (Table S5 in Multimedia Appendix 1). The reliance on provider-centric EHRs/EMRs may lead to difficulties in monitoring and assessment. Therefore, most studies (23/24, 96%) did not apply HISs in the entire process (NGRC data capture, intervention prescription, monitoring, and assessment). HISs were mainly used for (1) NGRC data capture (18/24, 75%), with limited patient-led (5/24, 21%) or system-led (1/24, 4%) cases; (2) intervention prescriptions (13/24, 54%); and (3) assessments (11/24, 46%) focusing on health care utilization, with minimal use in monitoring (2/24, 8%). Limited monitoring and a narrow focus on health care utilization data for assessments leave the sustainability of behavior change largely unaddressed. Given its critical role in improving long-term health outcomes, future research should aim to establish integrated frameworks that embed tethered or interconnected PHRs consistently across the 4 phases. Furthermore, the lack of details on HIS functionalities limits the ability to synthesize evidence and derive practical design recommendations for future interventions. Future studies should provide sufficient details to ensure reproducibility.

Although HISs have been used frequently in NGRC data capture, the dominance of active collection (eg, patient-administered or provider-administered surveys; 22/24, 92%) and the underuse of passive collection (eg, linkage with public datasets; 1/24, 4%) and GPS data (1/24, 4%) have left existing data underutilized, and active collection may cause user information to be unclear. On one hand, the majority of studies (17/24, 71%) relied on manual NGRC collection and documentation into EHRs/EMRs by stakeholders (Table S6 in Multimedia Appendix 1). Active data collection likely contributes to negative provider attitudes due to increased workload and limited consultation time. This potentially hinders the integration of NGRCs into routine care. Self-completed assessments using tethered or interconnected PHRs may address providers’ concerns, and both patients and providers hold positive attitudes toward them (Table S6 in Multimedia Appendix 1). However, tethered or interconnected PHRs were underutilized in the included studies. The gap between positive stakeholder attitudes and practice suggests missed opportunities to reduce provider burden and engage patients more actively in their care using PHRs. On the other hand, although the A-CHESS app achieved passive data collection, it relied solely on HISs as assessment tools and had limited provider access to behavioral data, which constrained the understanding of how HISs and providers influence behavior change and its sustainability. Similarly, the Envirome Web Service [77] passively collected NGRC data via public dataset linkages. However, the lack of detailed descriptions of the use of HISs in subsequent steps (eg, monitoring and assessment) further limited insights into their role in supporting sustained behavior change.

The complexity of provider types and roles in the entire process of NGRC-focused behavioral interventions (Table 1) further complicates implementation in clinical settings. Future studies should clearly define the responsibilities of every provider to enhance coordination across care teams, thereby improving intervention fidelity and consistency.

In summary, implementing tethered or interconnected PHRs throughout the intervention process, particularly in outpatient settings, is recommended (Figure 3). Connection with EHRs/EMRs can also achieve provider support. By ensuring privacy, security, and controlled access, PHRs can enhance user acceptance and usage [100]. Secure data exchange among patient apps, public datasets, sensors, and EMRs/EHRs can enable tailored interventions, timely issue detection, improved communication, greater engagement and adherence, better care coordination, higher satisfaction, and reduced treatment burden [86,101]. We have summarized 3 recommendations in Textbox 3.

The dominance of structured, measurable social needs interventions in the involved studies likely reflects that health care prioritizes short-term outcomes. In contrast, long-term lifestyle change interventions are overlooked. This may be due to operational challenges, including reimbursement constraints, lower adherence, long-term evaluation, etc [112-115]. However, lifestyle change interventions are crucial for preventing NCDs. To raise attention on lifestyle change interventions, incorporating NGRCs is beneficial for improving adherence and effectiveness. Future studies should focus on leveraging NGRC data to design and evaluate personalized, NGRC-focused lifestyle interventions aiming at advancing patient-centered and preventive care.

Current research has largely overlooked NGRCs beyond transportation barriers within the “destination accessibility” dimension of built environments. Neglecting other 4D elements, as well as social (eg, community crime, neighborhood safety, and socioeconomic data; included in 4 studies) and natural environments (specifically air pollution; included in 1 study), may limit our understanding of how specific NGRCs can inform behavioral interventions. Within the built environment domain, the exploration of “density,” “distance,” and “design” in the “5D” framework was limited, and “diversity” was not addressed. Future studies should systematically explore other less-studied NGRCs, such as social, natural, and other built NGRCs, to inform more holistic solutions of behavioral interventions using HISs.

Insufficient attention to objective, dynamic, and/or high-resolution NGRCs has left several key methodological challenges largely unaddressed. These include the modifiable areal unit problem, uncertain geographic context problem, neighborhood effect averaging problem, and problem of temporal variations in NGRCs (Figure 4) [60,116-120]. One likely reason is that 22 studies (92%) relied on actively collected, subjective, and static NGRC data, which require substantial participant effort and are therefore difficult to scale for capturing dynamic and/or high-resolution NGRCs. Passive data collection methods (eg, GPS tracking and linkage with public datasets) could help overcome these limitations by enabling more dynamic and/or high-resolution NGRC measurement. However, passive approaches alone cannot capture all real-time and/or high-resolution NGRCs, particularly those related to individual perceptions. Therefore, active data collection remains essential for assessing perceived NGRCs. While some studies have focused exclusively on perceived measures, others have underscored the added value of combining subjective and objective NGRCs to more comprehensively understand their associations with behavioral outcomes [121-124]. When collecting perceived dynamic and/or high-resolution NGRCs, issues, such as user vagueness and the quality of self-reported data, must be carefully considered. Future work should prioritize developing optimized and balanced approaches that integrate active and passive data collection, thereby leveraging dynamic and/or high-resolution NGRC measures to advance behavioral intervention research.

Furthermore, all involved studies illustrated the ongoing challenges of balancing granularity and comprehensiveness in assessing NGRCs. On the one hand, actively collected, subjective NGRC measures provide individualized insights but are limited in granularity and comprehensiveness. On the other hand, passively collected, objective indicators offer broad coverage but often lack individual insights and granularity, potentially due to existing dataset constraints. This trade-off underscores the need for integrated approaches that combine active and passive data to capture the complex, dynamic interplay between NGRCs and health behaviors.

To improve the granularity and comprehensiveness of NGRCs in various behavioral interventions, especially lifestyle change interventions, HISs—preferably tethered or interconnected PHRs in outpatient settings—should be considered. We have summarized 3 recommendations in Textbox 3.

Improved perceived value, convenience, patient autonomy, and privacy were consistently reported as outcomes of integrating technological (HISs) and content (NGRCs) innovations into behavioral interventions in all 10 included studies. On the other hand, social concerns persisted, including provider workload and the potential for discrimination. These benefits and concerns could be influenced by participants’ computer literacy, which was rarely reported in the studies. Given this gap, future research should explore how to balance technological and content innovations with unresolved social concerns, accounting for variations in computer literacy.

The NGRC information needs of providers, the clinical usefulness of NGRC data, and the usefulness of NGRC-related data visualizations are not yet well supported by existing evidence. Future studies should investigate which NGRCs should be presented in HISs, how they should be used in practice, and which interactive features best support their use.

Stakeholders’ insights into the integration of NGRCs and/or HISs into lifestyle interventions were not explored in the included studies. This highlights a critical gap in both technological and content innovations, as well as in addressing social concerns and in the integration of HISs and NGRCs into lifestyle change interventions.

We have summarized 5 recommendations focusing on enhancing NGRC and/or HIS integrations in behavioral interventions in Textbox 3.

We found limited coverage and inconsistent health outcomes of NGRC-focused behavioral interventions using HISs. Health outcomes focused mainly on health care utilization (eg, emergency visits), with limited coverage of clinical indicators, including medication adherence [97], asthma control [92], cancer screening [98], and blood pressure [87]. This limited coverage of clinical health outcomes limits the understanding of long-term behavior change and highlights the need for more comprehensive assessments. Moreover, the inconsistent health care utilization outcomes [88,91] may be caused by the inconsistent and/or insufficient reporting of 5 aspects. These aspects include (1) NGRC attributes (eg, definitions and spatiotemporal characteristics), (2) HIS types and functionalities, (3) assessment methods (eg, clinical metrics [eg, HbA_1c] and health care utilization), (4) stakeholders (eg, only HISs or various health providers), and (5) targeted diseases. Such heterogeneity causes low interstudy comparability and reproducibility. It also limits the ability to determine which behavioral interventions, leveraging specific NGRCs via particular HIS functions, are most effective, thereby complicating the translation of findings into evidence-based practice.

No studies have addressed key issues such as cost-effectiveness and ethical considerations. Future research should examine these critical factors to ensure successful implementation.

To develop evidence, we have summarized 3 recommendations in Textbox 3.

Our objective was to synthesize global literature on NGRC-focused behavioral interventions supported by HISs. However, we were unable to identify published, peer-reviewed studies conducted outside of the United States. Initiatives, such as the European Health Data Space (EHDS) [125] and China’s Healthy China Initiative [126], reflect growing efforts to integrate SDoH, including NGRCs, into HISs. On one hand, contextual differences do not preclude transferability. The contextual differences include cultural norms, socioeconomic conditions, and geographic contexts. With appropriate localization and system-level adaptation, the core principles identified in United States–based studies can be effectively applied in diverse settings. However, the limited diversity of study settings may obscure unique challenges and solutions relevant to low-resource or non-Western countries. Addressing these gaps is essential, and future research should ensure that context-specific NGRC-focused behavioral interventions using HISs are inclusive, adaptable, and globally relevant (Table 3). On the other hand, technological and political differences across the world do not influence the generalizability of our study. These differences include (1) HIS infrastructure, architecture, and integration capabilities, and (2) data standards, privacy regulations (eg, the European General Data Protection Regulation [127]), and governance policies. The commonalities of these differences help support the generalizability of the findings [128].

Our review has 4 key strengths. First, this article reviews the current state of NGRC-focused behavioral interventions using HISs. This provides a comprehensive analysis from 3 perspectives: technology, NGRC content, and stakeholders’ insights. Second, we searched a wide range of databases, leading to publications from diverse disciplines, including medical journals, environmental journals, and computer science conferences. Third, we consulted a library information specialist to develop the search strategy, ensuring its comprehensiveness and accuracy, which thus enhanced the research’s quality and credibility. Finally, to provide readers with sufficient context and allow them to assess the reliability of the evidence, we extracted and reported descriptive information, including study type, publication year, and sample size (Table S4 in Multimedia Appendix 1). We did not conduct a critical appraisal of the included studies, as it is not mandatory in scoping reviews. According to the Joanna Briggs Institute (JBI) methodology for scoping reviews [129], a critical appraisal is generally unnecessary unless it is directly relevant to the review objectives. Additionally, most quality assessment tools, including the Mixed Methods Appraisal Tool (MMAT) [130], primarily evaluate study design, data collection, and methodological rigor. Therefore, applying such tools would not provide meaningful information for our objectives (eg, HIS functionalities and NGRC indicators).

Our study also has several limitations that should be considered. First, NGRCs are ambiguous, suggesting that they are open to multiple interpretations or are vague. For example, in various studies, neighborhood safety may refer to different things, but there are no clear details of this concept in the studies. However, resolving this ambiguity is challenging because of the lack of standardized definitions and the context-dependent characteristics of NGRCs. Therefore, future studies should adopt detailed and consistent definitions of NGRCs, ideally guided by standardized terminologies or ontologies. These terminologies or ontologies can leverage and expand existing terminologies (eg, “community details” in Systematized Nomenclature of Medicine—Clinical Terms [SNOMED CT] and Logical Observation Identifiers Names and Codes [LOINC]) and SDoH ontologies [24].

HISs are often seen as passive recorders of health data. However, tethered or interconnected PHRs can actively participate in the entire process of NGRC-focused behavioral interventions, especially in outpatient lifestyle change interventions, with the relatively limited help of health care providers. For example, platforms like the iManageCancer platform for cancer patients and the iManageMyHealth app for their physicians provide personalized content from reliable public resources on cancer. However, these platforms do not currently consider NGRC factors and focus on behavioral interventions [3]. This underscores the importance for providers and policymakers to utilize tethered or interconnected PHRs to collect comprehensive, dynamic, and static NGRCs; deliver timely behavioral interventions in outpatient settings; monitor behaviors and the health status; and assess interventions. By doing so, they can achieve public health goals and promote health equity.

In this study, we found that most NGRC-focused behavioral interventions utilized transportation barriers to support social needs interventions. Meanwhile, HISs were involved only in specific aspects of the NGRC-focused behavioral intervention process. To achieve comprehensive dynamic and/or high-resolution NGRC considerations and the full integration of HISs throughout the behavioral intervention process, given the limited and heterogeneous evidence, further research is needed to determine how tethered or interconnected PHRs can effectively support NGRC-focused interventions, especially lifestyle change interventions. During implementation, it is essential to address social concerns alongside technological (eg, HIS integration) and content innovations (eg, NGRC integration). We compiled a list of 15 recommendations for future research. These recommendations are likely to accelerate the integration of tethered or interconnected PHRs in NGRC-focused behavioral interventions.