Shutdowns and personal restrictions introduced in March 2020 during early days of the COVID-19 pandemic required family medicine clinics to shift operations rapidly to virtual care in Ontario.1 Clinicians faced major logistical challenges in providing primary care using high-quality secure virtual care tools such as videoconferencing, real-time chat, private messaging, and file transfer. Physicians preferred to use tools that were readily available, with secure messaging frequently favoured over video visits.2 It was important for medical clinics to continue to provide patient care to reduce harms from unintended consequences of isolation, quarantine, and limited doctor-patient interactions.

A survey of patients conducted in Ontario in early 2021 by Ashcroft et al reported experiences with telephone and video visits as modalities used for virtual care, with most patient respondents indicating telephone and video visits reduced time and costs and were more convenient than in-person appointments.3 In this article we describe our experience with pioneering an innovative hybrid approach to providing virtual and in-person primary care services at our clinic (Get Well Clinic in North York, Ont) between March 2020 and May 2022 by restructuring administrative processes and designing a custom e-Platform website (https://www.getwellclinic.ca) using open-source and low-cost software. Installing and hosting our own e-Platform allowed us to pivot to virtual care rapidly—within a week in March 2020—which kept our clinic open to our patients during a critical period.

Building the e-Platform

We purchased an Internet service plan from a national provider that included 5 static Internet protocol addresses for approximately $150 per month. To secure the network from cyberattacks we initially used a DD-WRT (Dresden wireless router) firewall and later a pfSense firewall (no cost). We used virtualized server hardware (cost of approximately $1500, although this expense had already been absorbed as an existing operating cost to locally host our electronic medical record [EMR] system: Open Source Clinical Applications and Resources [OSCAR] version 19, Community Edition) running Linux Ubuntu long-term support with QEMU kernel-based virtual machine hypervisor (no cost) and using a virtual website server LAMP stack (ie, a software bundle of Linux, Apache, MySQL, and PHP). For publishing, editing, and changing content on our website we used Joomla Content Management System (no cost), with some low-cost Joomla plug-ins for live chat and video chat functions (JChatSocial Enterprise, approximately $75) and for online booking and appointment management (OS Services Booking, approximately $55). We eventually added a self-hosted Jitsi Meet server (no cost) for transport layer security and secure sockets layer encrypted peer-to-peer WebRTC-based videoconferencing. The entire e-Platform was put together with our physician information technology expert (K.L.) and volunteers (in-kind support).

Training and implementation

We retrained clinic staff and volunteers for a new process for patient check-in and aftercare. Volunteers created online videos to teach patients how to register for and access our e-Platform. Clinic staff verified patient information and confirmed their identity with a telephone or video call before activating their account. Once their accounts were activated, patients could use a live chat feature to converse with front-desk staff. Health card validation was done via a video call for visual confirmation of the patient’s identity and physical health card. After check-in health care professionals could join the chat or live video call and could invite interdisciplinary professionals (eg, dietitians, psychotherapists, translators) or a patient’s family member in another location to join the video call seamlessly. For patients without a Web camera, virtual visits could be downgraded to audio only with the chat option still available for file transfers of patient records, images, or requisition forms. Providers who wished to use the Ontario Telemedicine Network (OTN) could still create an OTN link on the OTN website (https://otnhub.ca) and then paste the OTN link into the chat for the patient to access.

Pivots and shifts

We were able to conduct about 75% of patient care visits as video-enabled virtual appointments with our e-Platform early in the pandemic, reaching close to 85% in the first few weeks following lockdown in spring 2020 (Figure 1). High-quality video-enabled encounters allowed us to provide better diagnoses (eg, allowing a clinician to differentiate a perianal abscess from a benign hemorrhoid) and better mental health care (eg, by seeing patients during psychotherapy without masks). By December 2021 telephone calls surpassed video chats as the dominant mode of virtual care, and by May 2022 video-enabled visits dropped to below 10% of all virtual appointments (Figure 2) in tandem with an increased return to in-person care (Figure 3).

Figure 1.

Percentage of patient care visits conducted virtually per month at the Get Well Clinic in North York, Ont, between January 2020 and May 2022

Figure 2.

Percentage of virtual visits conducted by telephone or video (e-Platform) per month at the Get Well Clinic in North York, Ont, between January 2020 and May 2022

Figure 3.

Patient care visits per month at the Get Well Clinic in North York, Ont, between January 2020 and May 2022: Total number of appointments, number of virtual appointments, and number of virtual appointments conducted by video.

Strengths and weaknesses

By enabling virtual care at Get Well Clinic we were able to adopt a hybrid virtual and in-person approach to provide primary care safely during a global pandemic. We used virtual care where clinically appropriate and allowed for in-person care as needed. Use of open-source tools allowed for rapid implementation of this approach at low financial cost. However, use of virtual care technology did not address the problem of higher administrative burdens associated with providing a feature-rich video-enabled virtual care platform.

Because the COVID-19 pandemic was a novel event, there were no known implementations of similar approaches (prior to March 2020) for providing hybrid virtual care in the primary care setting during a global pandemic. We present a strategy that was tailored to our clinic’s circumstance and priorities, and we recognize this uniqueness constitutes an inherent weakness in terms of generalizability of our approach and results.

We chose to implement direct video visual confirmation of the patient’s identity at each visit to ensure the same level of identity management as in-person visits. This process contributed to increased administrative costs of providing virtual care. The strength of our results was due to priorities we applied when implementing technology. We operated a managed office clinic and could uniformly deploy change management and train all our care providers. To satisfy concerns regarding data privacy and ownership, we hosted our e-Platform website on our own local servers within our clinic with standard transport layer security and secure sockets layer encryption protocols, minimizing risk for data breaches by third-party vendors. We looked to open-source software for community-reviewed source code, essentially a type of peer-reviewed literature. We mitigated the risk of intruders randomly guessing video “room numbers” (ie, “Zoom bombing”) by using Jitsi Meet video links that were composed of complex alphanumeric URLs (website addresses) based on a cryptographic hash. Use of open-source tools was substantially less expensive for us in the long term than purchasing a commercially available product.

Online live chat with our front-desk staff allowed our clinic to communicate with patients in real time and allowed for multitasking by our limited number of staff. Live chat provided an alternative channel of communication to the telephone line, which could be busy or have missed calls. File transfer via the chat platform allowed us to protect patient privacy, security, and safety by avoiding the use of email and by not requiring patients to go to the clinic during the height of the pandemic.

Data from the previously mentioned 2021 survey of patients in Ontario by Ashcroft et al suggested most patients were keen to continue having telephone and video visits after the pandemic,3 but our data showed a waning use of virtual care later in the pandemic as restrictions eased. This might be explained by differences in practice population and approaches. Our study involved both rostered and unattached patients, whereas other studies primarily involved pre-existing doctor-patient relationships. We developed our own virtual care platform from open-source tools, while other studies used commercially purchased products that might have been integrated with the providers’ EMRs.1-3 The provincially supported OTN hub did not have file transfer or real-time chat capabilities at the time of our study. Providers in other studies primarily used asynchronous messaging (email, text, or secure messaging) and telephone calls for virtual care.2-4 It is not clear whether they implemented visual identity management systems or strategies to confirm patient identity at all visits. We have heard anecdotally that many providers implicitly trusted the identity of patients in virtual encounters based on patients’ confirmed telephone numbers or email addresses and based on providers’ familiarity with patients’ voices; however, these approaches cannot be relied on in settings with unattached patients (eg, walk-in clinic) or given potential risks of artificial intelligence deepfake manipulations. We deployed a consistent change management system for all our providers and staff, resulting in the initial predominance of video-enabled virtual care at our clinic in contrast with other studies in which telephone calls and asynchronous messaging were primarily used.2,3 Our project demonstrated what was required to achieve the next level of virtual care (ie, video enabled) compared with now commonly used telephone and asynchronous messaging. Our findings showed how our clinic gradually shifted over time from enthusiastic video use to a less burdensome mode of remote care—telephone calls.

Continuation of our hybrid approach presented substantial challenges. Virtual care required extra clinic resources to provide technical support and training for staff and patients. Adding another channel of communication with live chat required additional staffing, which also burdened our staff with managing multiple simultaneous live chats. These increased administrative duties and indirect financial costs were unsustainable given the state of public primary care funding, labour shortages, and rising salary costs. Virtual care options have matured since the pandemic, giving rise to several commercial products that integrate with EMRs for additional fees. It is our opinion that technology alone has not addressed administrative costs to make a worthwhile business case for using video-enabled encounters routinely in comprehensive primary care. However, we believe telephone calls and asynchronous messaging (email or private messaging) will continue to be viable tools in virtual care. Our open-source solution served a specific purpose during a particular moment in pandemic history. Given the benefit of hindsight, we would still opt to implement our own open-source solution rather than purchase a commercially available product.

Conclusion

Technology is not a panacea for solving problems with sustainable primary care today. We recommend that health care professionals, organizations, and policymakers consider administrative costs associated with developing, implementing, and maintaining virtual care solutions. Funding should be provided not only for purchasing commercial products but also for covering costs of clinic administrative support. Lower-cost patient-identity management solutions are also an area that must be addressed for virtual care encounters. This is an important requirement when developing virtual care solutions for unattached patients in remote areas who are physically unable to travel to clinics. In light of funding restrictions in Canadian primary care, we propose the use of community-developed open-source tools for robust, customizable, and cost-effective options for virtual care in family medicine.

Our clinic was able to adopt a hybrid approach to provide primary care safely during a global pandemic, offering patients both virtual and in-person visits. Our novel approach provided unique insight on the use of open-source tools for virtual care in family medicine during a pandemic. Sustainability of feature-rich video-enabled virtual care in primary care settings requires more exploration and support.

Notes

We encourage readers to share some of their practice experience: the neat little tricks that solve difficult clinical situations. Praxis articles can be submitted online at http://mc.manuscriptcentral.com/cfp or through the CFP website (https://www.cfp.ca) under “Authors and Reviewers.”

Footnotes

Competing interests

None declared

Copyright © 2024 the College of Family Physicians of Canada