The opioid crisis remains a significant public health challenge [1], highlighted by the rising number of deaths worldwide [2-4]. In North America, illicit drug poisoning deaths can be attributed to a contaminated drug supply saturated with synthetic opioids, particularly fentanyl [5]. Simultaneously, opioid use disorder (OUD) is on the rise [6], with 21.4 million cases worldwide in 2019 [7]. Effective medications for OUD, such as opioid agonist therapies (OATs), reduce illicit drug poisoning deaths [8,9]. Given that primary care services are often at the forefront of diagnosing and treating OUD, evidence shows that accessible primary care services can help achieve better health outcomes for people with OUD [10,11].

The COVID-19 pandemic introduced new barriers to traditional approaches to OUD care, including in-person primary care services, due to the need for social distancing and reduced in-person interactions [12,13]. In response to COVID-19 public health measures, health care systems implemented remote care solutions, transforming the delivery of health care services [14] and accelerating the adoption and use of virtual health care [15-17]. During this same period, much of the progress that had been made to address the opioid crisis was reversed [18-20]; illicit drug poisoning deaths increased as a result of limited health services [21], poisoned supply [22], and solitary drug use [23]. In the context of these challenges, and as a result of pandemic-related public health measures, virtual care emerged as a potential solution to address the unique needs of people with OUD by providing essential health care services remotely [24].

The widespread shift to virtual care facilitated continuous access to care during the pandemic. This shift may have been helpful for individuals with OUD [25,26], many of whom experience barriers to health care due to intersecting effects of criminalization, discrimination, and social marginalization [13]. Virtual care allows clinicians to deliver essential health care services to people with OUD [27,28], offering improved accessibility [29] and the ability to deliver OAT [30] while reducing barriers related to transportation, mobility limitations, and stigma [31]. However, virtual visits may limit clinicians’ ability to fully assess patients’ physical and emotional well-being [27] and may introduce technological barriers for those with limited internet access and low digital literacy [32,33]. In addition, virtual care may not be suitable for all aspects of OUD care as some interventions, such as injectable OAT [34], require regular in-person management [27]. The integration of virtual care into the management of OUD within the primary care context may improve health care delivery, but its limitations, which make in-person visits necessary, must be acknowledged to maximize its efficacy.

Virtual platforms have become increasingly common for primary care services—a trend that is expected to persist beyond the pandemic [35]. Therefore, it is essential to understand the benefits and limitations of current virtual care interventions to shape future health care delivery models and optimize the care provided to patients with social and clinical complexities. This scoping review aimed to synthesize the current knowledge landscape and research gaps and offer insights into the potential of virtual primary care for people with OUD.

This review informs an ongoing mixed methods study investigating changes to primary care access and patient outcomes following the rapid introduction of virtual care for people with OUD in British Columbia [36]. We conducted the review in accordance with the Joanna Briggs Institute Reviewer’s Manual [37,38] and the framework suggested by Arksey and O’Malley [39]. We registered our protocol on the Open Science Framework [40] and report our findings using the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) [41] and the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) for Abstracts checklists [42] (Multimedia Appendix 1).

We established our eligibility criteria a priori and present these in Textbox 1. We included studies that focused on the concepts of “virtual care,” “primary care,” and “people with OUD” and excluded studies that did not include all 3 concepts. We defined “primary care” as “the provision of integrated, accessible health care services by physicians and their health care teams who are accountable for addressing a large majority of personal health care needs, developing a sustained partnership with patients, and practicing in the context of family and community” [43]. Therefore, primary care settings are those where patients could receive comprehensive care (ie, treatment for any type of health issue) [44], including physician’s offices, community health centers, and specific outpatient clinics. We included some settings that may not traditionally be considered primary care—such as certain opioid treatment programs, syringe service programs, or veterans’ health association programs and clinics—when there was evidence that the setting also attended to health care issues beyond OUD [45-47].

We developed search strategies [48-52] by combining Medical Subject Heading terms (eg, analgesics, opioids, telemedicine, and primary health care) and free-text keywords (eg, methadone, virtual, and family doctor) using Boolean (eg, AND, OR, and NOT), truncation, and wildcard operators. Librarians from the University of British Columbia and Simon Fraser University verified the search strategies. Our final search strategy is included in Multimedia Appendix 2. From December 6, 2022, to December 8, 2022, one reviewer (SN) systematically searched MEDLINE (Ovid) [49], CINAHL Complete (EBSCOhost) [51], Embase (Ovid) [50], and the Web of Science Core Collection [48] with no end date. From May 2023 to June 2023, we searched gray literature using the Canada Commons database [52], Google Scholar, Trip, Grey Matters, and select organization websites (ie, Canada Health Infoway, British Columbia Centre on Substance Use, Centre for Addiction and Mental Health, and Canadian Research Initiative in Substance Misuse). Following the search and extraction, the bibliographies of the included studies were systematically checked for additional references [53].

One reviewer (SN) identified and removed all duplicates using Covidence (Veritas Health Innovation) [54], a systematic review software, after which 2 reviewers (SN and EG) screened the identified studies in a 2-stage screening process (title and abstract screening and full-text screening). We achieved a 75% minimum agreement in a pilot test of our screening process and reviewed any conflicts. During screening, some studies did not have well-defined care settings, leaving raters to apply the inclusion criteria differently. Furthermore, the studies spanned a range of methodologies and outcomes, introducing additional challenges in achieving consistent ratings. We reconciled discrepancies at both stages through discussion with the research team and recorded the reasons for study exclusion at each step.

We (EG, LH, RKM, and SN) created a data extraction chart (Multimedia Appendix 3) to capture relevant information (ie, setting, benefits, challenges, virtual modality, and patient population) in Covidence. To ensure its efficacy, we first piloted this chart on a small sample of studies. On the basis of insights from this pilot test, we refined the chart to better capture the nuances of health service features specific to primary care settings. This revised chart became the cornerstone for the subsequent data extraction from the studies. In parallel, EG and SN conducted quality appraisals of the studies using the Quality Assessment With Diverse Studies (QuADS) tool [55]. The QuADS tool assesses the quality of the methods, evidence, and reporting of studies included in systematic reviews. Our appraisal process involved scoring the studies on a scale from 0 to 3, where the score is derived from an evaluation across several areas, including the clarity of the research aims and robustness of the study design (Multimedia Appendix 4). No study that reached the data extraction stage was excluded due to their QuADS score. Following data extraction and quality appraisal, SN and EG reviewed the data to reach a consensus. We transferred the harmonized data to Microsoft Excel (Microsoft Corp) to conduct inductive content analysis, which led us to identify patterns, which we subsequently organized into themes.

For theme categorization, we used the Consolidated Framework for Implementation Research (CFIR) [56], a widely recognized meta-theoretical model previously applied in other virtual care reviews [57-59]. The CFIR comprises five major domains (Figure 1):

We discuss facilitators of and barriers to the implementation process (domain 5) within each of the other 4 domains.

We identified a total of 1474 studies from our searches of the Embase (Ovid; n=780, 52.92%), Web of Science Core Collection (n=283, 19.2%), Ovid MEDLINE (n=263, 17.84%), and CINAHL Complete (EBSCOhost; n=142, 9.63%) databases and through citation searching (n=6, 0.41%). After removing 36.36% (536/1474) of the identified sources due to duplication, we screened the titles and abstracts of 938 studies, retaining 155 (16.5%) for full-text review. After a double review process, we identified 28 papers for data extraction and quality assessment. A PRISMA flow diagram details the exclusions (Figure 2). In this section, numbers within parentheses indicate the frequency and percentage of the studies included in the review that support the statements.

Table 1 provides a summary of the characteristics of each study with its corresponding citation. Most studies (25/28, 89%) were based in the United States and were published after the emergence of COVID-19. Only 14% (4/28) of the studies were conducted before the pandemic [60-63], emphasizing the paucity of research on virtual primary care for people with OUD conducted before 2020. Original research constituted the largest group of studies (11/28, 39%), using various study designs. Most studies had either quantitative (15/28, 54%) or qualitative (4/28, 14%) designs, whereas 4% (1/28) of the studies had mixed methods designs. Other studies reported their findings as commentaries (7/28, 25%) or economic evaluations (1/28, 4%).

The key objectives and results identified in each study are summarized in Table 2. A total of 68% (19/28) of the studies identified incorporating virtual care as part of their primary study purpose or design, whereas the remaining studies (9/28, 32%) included virtual care as an ancillary component of their research. The studies commonly reported the introduction of virtual modalities due to COVID-19 (17/28, 61%), with a few of these papers (6/28, 21%) also examining changes to or increases in existing virtual modality use in response to the pandemic. Outcomes included reports of shifts in health service use, such as changes in appointment modalities (16/28, 57%), frequency or demand (4/28, 14%), and differences in prescription duration (4/28, 14%). We reviewed the number of articles that reported on demographic factors given our interest in equitable practices and understanding disparities in research and health care delivery. Given that it is uncommon for commentaries to report demographics, we excluded them in the following analysis. Of the remaining 22 articles, 11 (50%) reported on sex or gender, 13 (59%) reported on race or ethnicity, and 10 (46%) reported the age of the participants. A total of 18% (4/22) of the studies specified collecting data on participants’ gender [63,64,74,81], whereas 23% (5/22) specified the participants’ sex [36,62,76,77,80]. In total, 9% (2/22) of the studies were unclear on whether they were reporting on sex or gender (eg, female/male were used, but the study did not specify whether sex or gender were reported) [60,82]. Benefits and challenges of using virtual primary care for people with OUD are grouped within the CFIR domains in the following sections and in Table 3.

Primary care (including but not limited to OAT initiation and management) was often delivered through telephone appointments (18/28, 64%), with fewer papers focusing on video modalities (14/28, 50%) or not specifying the type of synchronous virtual visits (9/28, 32%; Table 4). Some studies (7/28, 25%) emphasized the importance of providing multiple modality options, such as virtual and in-person appointments (ie, hybrid).

Where studies described primary care prescription of OAT, most (16/28, 57%) referred to provision generally without specific mention of the stage or length of care. Others described OAT initiation (13/28, 46%) or follow-up and stabilization (9/28, 32%) for management of OUD virtually. Outside of providing OAT, mental health care was most commonly provided (14/28, 50%), followed by treatments for sexually transmitted and blood-borne infections (6/28, 21%) [68]. In addition, harm reduction services were provided (such as syringe exchange and naloxone distribution with the support of phone consultations and home deliveries [80]), although to a lesser extent (3/28, 11%). Few studies (6/28, 21%) described other types of primary care provided (eg, on-site vaccinations [64] and in-person diabetes management [73]) and lacked specific details on the care provided (eg, it was listed as “other health needs” [63] or “other clinics” [62]).

The most common facilitator described in the included studies was regulatory changes (11/28, 39%) to support the delivery of virtual primary care for OUD, such as the waiver of the Ryan Haight Act in the United States, which allowed clinicians to prescribe buprenorphine without the previously required first in-person appointment [77]. Another common facilitator was COVID-19 public health measures (4/28, 14%). Due to the increased COVID-19 risk to patients and clinicians associated with in-person appointments, people with OUD could use virtual primary care to adhere to social distancing measures and reduce their COVID-19 exposure risk [84]. The reduction of COVID-19 exposure risk was also described as a benefit of virtual care.

Regulatory barriers were less commonly described (7/28, 25%) but included uncertainty in the permanence of the COVID-19 regulatory changes [72] and limited flexibility in providing methadone via virtual modalities [80]. Billing structures were described as both facilitators and challenges. For example, new billing structures or fee codes introduced during the COVID-19 pandemic (5/28, 18%) facilitated virtual care delivery. However, some clinicians experienced challenges with billing (5/28, 18%), stating that payment models did not equally compensate virtual and in-person care [65] or that virtual engagement was not included as a billable service [63]. Insurance policies yielded negative perspectives (4/28, 14%), where inconsistent [71] or insufficient [69,78] coverage of virtual care and challenges with reimbursement [86] were reported. Urine drug screenings were mentioned in 39% (11/28) of the articles [60,62,63,66,70-72,74,80,85,86]. Some articles described that changes to federal mandates during the COVID-19 pandemic reduced (3/28, 11%) [66,80,86] and facilitated innovative approaches to urine drug screens (3/28, 11%) [70-72], such as take-home kits or remote viewing.

Several studies (6/28, 21%) described clinical resources or guidelines as essential factors for facilitating virtual primary care implementation, ensuring that clinicians and patients could navigate virtual appointments. Such guidelines ranged from documents produced by the Substance Abuse and Mental Health Services Administration to guide clinics through policy changes to providing OAT virtually [70] or individual clinic protocols for virtual delivery of OUD care [76]. Other resources included the provision of communication devices to clinicians and patients (5/28, 18%) and training or education for patients to prepare for their virtual appointments (5/28, 18%). A total of 14% (4/28) of the studies noted that certain primary care clinics were better positioned to adopt virtual primary care. For example, clinics in urban or community-based settings [74] or larger clinics [77] were more likely to offer virtual OAT induction, although the reasoning behind these differences was not explained in the research.

Challenges affecting implementation at the clinic level included the need for more administrative support for physicians (4/28, 14%) and a lack of guidelines and resources (4/28, 14%) to support clinicians in providing evidence-based care (eg, lack of protocols for virtual primary care [85]). The studies also highlighted the challenges associated with specific clinical settings (4/28, 14%), noting that accessing OUD care without an in-person examination was more difficult in primary care solo practices versus specialty substance use facilities, opioid treatment programs, community clinics, and federally managed health administrations [74]. Furthermore, primary care practices in rural settings were less likely to provide virtual primary care services, possibly due to the lower demand for virtual services as COVID-19 risk was lower than in urban areas [74]. A few studies (3/28, 11%) described low demand for virtual care in general from both patients and clinicians [71,74,81], resulting in slower uptake in some clinics.

The most common benefit for patients was the high accessibility of virtual care (13/28, 46%), particularly for rural patients (6/28, 21%). This advantage was linked to other benefits, such as alleviating travel challenges (10/28, 36%) and flexibility of appointments (4/28, 14%). Factors such as confidence in virtual care (2/28, 7%), its perceived convenience and efficiency (2/28, 7%), improved continuity of care (3/28, 11%), and the ability of virtual modalities to address patients’ stigma concerns (4/28, 14%) also emerged as pivotal factors. In addition, younger age (1/28, 4%) and cost savings (2/28, 7%) positively influenced patients’ engagement with virtual primary care.

The gap between people who can easily use and access technology and those who cannot (ie, “the digital divide”) remains the primary obstacle to the uptake of virtual care for people with OUD (12/28, 43%). This is compounded by a lack of technological knowledge or equipment (6/28, 21%), mistrust of virtual care (3/28, 11%), and absence of a private or secure space for appointments (2/28, 7%). Intersectionality between specific marginalizing patient attributes, such as being an older adult (1/28, 4%) and racialization (3/28, 11%), resulted in inequity in virtual primary care access due to structural barriers of the health care system.

Clinicians found virtual care beneficial, particularly when managing stable patients (5/28, 18%) and patients who could attend appointments consistently (5/28, 18%). Clinicians also found virtual care beneficial for continuity of care (3/28, 11%). Furthermore, previous experience with OAT or using virtual modalities (3/28, 11%) and a genuine interest in delivering virtual primary care (2/28, 7%) facilitated its adoption among clinicians.

Conversely, apprehension about liability or heightened risks associated with offering virtual primary care for people with OUD (10/28, 36%) were seen as challenges among clinicians. Concerns also emerged over the adequacy of phone-based relative to video-based appointments (3/28, 11%), challenges of fostering a therapeutic relationship through virtual visits (3/28, 11%), and the belief that in-person care was necessary for patients (2/28, 7%). Some clinicians also reported a preference for in-person care (3/28, 11%). Additional barriers included technological challenges (3/28, 11%) such as low digital literacy and inadequate technology equipment.

Over a quarter of the studies (8/28, 29%), most of which (7/8, 88%) were original research, received high scores through the QuADS tool. In contrast, commentaries (5/28, 18%) and conference abstracts (5/28, 18%) predominantly received lower scores. The scores obtained through our quality assessment using the QuADS tool are grouped by study type (Table 5).

In reviewing 28 studies on the experiences of people with OUD and primary care clinicians with virtual modalities in primary care, we sought to explore the current state of the literature, identify research gaps, and provide insights into the potential benefits and challenges of virtual approaches. The CFIR constructs were instrumental in our investigation, guiding us in identifying the organizational-, structural-, and individual-level factors that influence the implementation and ongoing use of virtual primary care for people with OUD. Our findings underscore that, with careful consideration of the implementation factors at all levels, virtual modalities in primary care hold promise as a feasible, acceptable, and effective option for people with OUD as complements to in-person primary care, aligning with research in other settings [88-90].

Our review highlights that research related to virtual primary care for the management of OUD is limited and even more so for the broad range of primary care provided for people with OUD beyond OAT despite the rapid increase in the availability of virtual care resulting from the COVID-19 pandemic. The available details on the range of primary care services offered virtually were scarce, but they did include areas such as mental health care, sexual health care, and harm reduction services. For example, people with OUD are known to frequently experience comorbid conditions (eg, HIV/AIDS, hepatitis C, and hepatitis B [91]), infections (eg, endocarditis and septic arthritis [92]), and other frequent or episodic health care needs (eg, wound care [93]) that require timely and ongoing primary care. However, the extant literature on virtual primary care for people with OUD focuses on management of OUD and not on other primary care needs, and few of the included studies (6/28, 21%) reported how or whether these conditions were addressed. This remains a core knowledge gap in the literature.

Relying on virtual modalities to increase access to health care was a common finding [94]. Virtual visits remove the necessity for physical travel to medical appointments and reduce associated costs (eg, transit or gas, taking time off work, and childcare). By offering the convenience of accessing care from a smartphone or other communication devices, virtual care can also support continuous access to and retention in OAT [95]. Patient lifestyle and stability play a significant role in driving their use of virtual care. Individuals with childcare responsibilities [79,85], limited flexibility during work hours [65,79], or those who are more stable on OAT [65,78,86] may find virtual care well-suited to meeting their care needs, while minimizing disruptions to their day-to-day lives.

The digital divide, the gap between those who can use and access technology easily and those who cannot, emerged as a central theme in multiple studies (12/28, 43%). Disparities in the components of the digital divide—such as access to technology, internet connectivity, and digital literacy—can affect the ability of people with OUD to use virtual care and its effectiveness. Furthermore, clinicians’ belief that virtual primary care was unsuitable for some patients, too risky due to patients’ health status, or associated with high liability also hindered its adoption, aligning with concerns found in other (ie, non–primary care) clinical settings [96,97]. With the challenges associated with patient stability and social determinants of health (eg, digital technology access, longevity of OAT treatment, and housing), providing care to people with OUD using solely virtual modalities may unintentionally limit equitable health care access. However, clinicians who believe that virtual primary care is unsuitable for unstable patients may be ignoring patient preferences [70]. While virtual primary care for less stable patients is a nuanced topic and needs concerted research efforts and practice approaches, evidence is emerging that virtual primary care can provide a safe supplement to in-person visits for unstable or new patients [72,78,98]. To enhance the effectiveness of virtual primary care for this population, offering diverse appointment types can accommodate patient preferences, socioeconomic factors, and health circumstances, fostering patient autonomy and promoting patient-centered care.

Virtual primary care for people with OUD is promising and feasible. However, this review also highlights that there may be disparities based on patient age [62,85]; language [85]; race or ethnicity [62,77,85]; sex [62]; and experience of the digital divide [65,68,70-72,76,77,80,81,84-86] due to rurality, socioeconomic status, or limited access to technology. These disparities pose challenges for implementing and accessing virtual care [99]. Without deliberate efforts to address these access barriers, virtual primary care could perpetuate inequities and worsen existing disparities in OAT access [100-102].

There are several limitations to this scoping review. Scoping reviews inherently possess limitations in terms of depth of analysis and generalizability [39], and the possibility of missing studies needs to be acknowledged. This is attributable to multiple factors; among them is our stringent inclusion criteria, which not only resulted in excluding studies with unclear or inadequately described settings but also affected our interrater reliability score. Furthermore, our definition of virtual visits excluded asynchronous forms, and we did not consider care settings that provided only OAT, which excludes the recent growth of literature on digital services for treatment of OUD as they take place within stand-alone addiction treatment services. This criterion was in place to align with our broader research project [36] and maintain a manageable review size. We acknowledge that there are valuable lessons from virtual care models provided for other populations and conditions, which may enrich our understanding and approach in this area but were not captured in this review. In addition, studies published in languages other than English were not captured. There was also a variance in the quality of the included studies. Some of the included studies (12/28, 43%) received low QuADS scores due to limitations in study information (eg, conference abstracts with restricted word counts). Furthermore, despite the growing demand for research that distinguishes between experiences associated with sex and gender [103,104], many of the studies we reviewed appear to conflate these distinct experiences (eg, reporting participant gender as female or male) or did not report any sex- or gender-based analyses. In addition, across the studies we reviewed, the heterogeneous research methods complicated data comparison and aggregation. Most studies (25/28, 89%) were conducted in the United States, which can limit or challenge generalizability to other health system contexts. For example, regulations and billing, part of the CFIR’s outer setting domain, differ dramatically across settings [105-107]. While qualified physicians in the United States and Canada can prescribe buprenorphine, the United States previously required a waiver under the Drug Addiction Treatment Act, whereas there is no such requirement in Canada [108]. Similarly, methadone treatment requires access to specialized clinics (ie, opioid treatment programs) in the United States [109], but in Canada, it can be prescribed in various health care settings [110]. Nonetheless, this scoping review presents novel results as the first of its kind covering virtual primary care for people with OUD. Our methodologically rigorous approach and the quality appraisal of the included studies make this review a valuable contribution to the literature given the sparse research in this field.

This scoping review highlights the promise of virtual primary care as a feasible, acceptable, and effective option for people with OUD, particularly in enhancing accessibility and continuous access to care. However, addressing the digital divide and mitigating disparities based on demographic and socioeconomic factors will be critical in realizing the full potential of virtual care. Future research and policy initiatives should prioritize efforts to bridge these gaps, ensuring that virtual primary care becomes an inclusive and equitable platform for delivering comprehensive care to and for people with OUD. In addition, there is a need for future research to explore the use of virtual modalities for primary care needs beyond OAT management to fully understand and enhance virtual care for people with OUD.