Wound care is a critical component of comprehensive home health services. The prevalence of older adults as recipients of home care is notable, with a substantial proportion necessitating wound care due to conditions such as pressure injuries, postsurgical recovery, and complications arising from venous ulcers. Older adults prefer to receive wound care in community settings [35], citing pain, transportation barriers, and risk of exposure to infection in the clinic as primary drivers of preference [36]. Patients and family caregivers frequently report feeling inadequately prepared to participate effectively and confidently in wound care [37]. The presence of nonhealing wounds and associated complications can profoundly affect quality of life, increasing risk of infections and hospitalization [38].    

In the Canadian context, the potential for an AI-enhanced DWCT to increase access to best practice care is particularly significant.  The “Caring for Canadians” study highlights a significant gap between the supply and demand for HHR in Canada, a gap that is already evident and expected to widen over the next decade, with rural and remote communities feeling the impacts most acutely [39]. Specialist HHR, such as NSWOCs or additional training in wound care, is in high demand. However, recruiting and retaining specialized health care workers is challenging, which has led to the adoption of innovative, technology-enhanced care models to optimize resource use [2,24].

In Canada, geographic location is a primary factor determining access to care, with isolated and remote communities experiencing the highest levels of deprivation [40]. These communities are frequently located in northern regions of the provinces and territories and are often home to Indigenous groups [40]. First Nations, Métis, and Inuit people, along with those of African, Caribbean, Southeast Asian, and Latin descent, have higher prevalence and incidence of diabetes and higher rates of lower extremity amputation because of diabetic foot ulcers and peripheral arterial disease [41]. This is particularly evident in Indigenous populations, where the rate of major amputations due to diabetic wounds is up to 49 times higher than in non-Indigenous populations [42]. This may be attributable to poor access to health care, barriers to clear communication, cultural differences, and discrimination [43]. Within this context, the AI-augmented DWCT offers an opportunity to increase access to best practice care by leveraging available resources and supporting available staff to work at the frontiers of their scope of practice (Textbox 1). 

In the Giishkaandago’Ikwe Health Services case study, the DWCT enabled a PSW who had direct contact with the patient to complete a wound care assessment even though this typically would be conducted by nursing staff. The DWCT tool enabled the team member who was geographically co-located to effectively capture the data needed for the wound care team to create a full care plan, without requiring the patient or the specialist providers to travel.  In such scenarios, AI-augmented DWCT can support an available staff member to work at the frontiers of the scope of practice while maintaining direct connections for clinical escalation and consultation when needed.   

 This is particularly relevant in the Canadian rural and remote context, where geography and inclement weather can act as a barrier to access. A 2020 review of barriers to equitable access identified the lack of year-round roads, extreme winter weather events, and the absence of rural and remote hospital staffing as key contributors to poor access to care in Indigenous communities [44]. That same review identified telehealth and telehealth-enabled consultation models as a mitigator of inequitable access. Furthermore, the review noted that telehealth enables community health workers and nurses to provide additional services in their local community and reduces the burden on patients to travel long distances to access care. This has important implications for equitable access, as community care providers typically have higher levels of trust and greater ability to provide culturally safe care [45].

The DWCT augments traditional “telehealth” in wound care by adding empirical data in the form of high-quality, time-lapsed photos in a “dashboard” that can be accessed by patients, caregivers, and the entire wound care team. This enables tracking of wound trajectory and gives contextual clinical data to geographically remote specialist providers such as NSWOCS, endocrinologists, and vascular surgeons.  Additionally, the dashboard supports team communication as it offers a shared point of reference to facilitate collaborative care planning.  In the context of an optimized scope of practice, this feature provides an additional layer of protection—ensuring specialized clinical oversight of geographically remote patients being cared for by practitioners working at the frontier of scope of practice.  Furthermore, this supports staff who may not have clinical training in wound care by providing avenues for consultation and escalation to complement their clinical judgment.  Ultimately, this benefits patients who receive care from a trusted provider in their community, while retaining the benefit of review and consultation from clinical experts.  

The AI-augmented features of the DWCT expand access to best practice care by (1) enabling patients in remote communities to access services without necessitating travel, (2) empowering clinicians practicing in rural and remote communities to work with an optimized scope of practice, and (3) supporting communication and ensuring all members of the care team have access to relevant clinical information. 

 Optimization of clinical resources is required to ensure efficiency and maximize the number of patients who receive best practice care. The AI-enhanced DWCT tool can increase clinical efficiency by conducting rapid assessment of wounds and streamlining consultations and escalations [28]. A key component of effective wound care is assessment of wound size. Traditionally, this is done with paper rulers, and the wound size is calculated using a formula that multiplies length by width. However, this traditional method is ill-suited to irregularly shaped wounds, leading to overestimation of size in 36%‐74% of nonrectangular wounds [46-48]. Patients with darker skin tones, wounds characterized by diffuse edges, irregular shapes, necrotic tissue, or unhealthy surrounding tissue are more prone to having their wounds overestimated in size [46]. A more accurate calculation of wound size can be captured using a tracing method, where transparency film is placed over the wound and outlined, but this is reported to be time-consuming [49].  Accurate and objective measurement of wounds is essential as a change in wound size is a key predictor of wound healing trajectory [5]. Digital wound measurements powered by AI are emerging as a more efficient and accurate means to measure wounds [46,50-52] (Textbox 2).

This case study highlights the benefits of tools that enable the clinical care team to optimize their time and streamline processes for escalating concerns to specialized care resources. Use of the DWCT is faster and clinically equivalent in accuracy compared to manual methods of wound measurement and capture [29,53]. Enhanced DWCT reduces the overall time required for wound measurement and documentation and has been shown to result in an average time saving of 1.01‐2.39 minutes per wound assessment when compared to the traditional manual methods [29]. Given the known constraints on clinicians’ time, burden of documentation, and anticipated workforce shortages, these time savings represent an opportunity to increase both efficiency and access.  

AI tools can also help identify and escalate wounds that require review.  HealingIndex is a new feature of the DWCT that uses machine learning algorithms to assess the photographs of the wounds that are submitted by clinicians. This tool analyzes the photo to identify wound characteristics associated with healing including wound size, tissue composition, and exudates, to predict the healing trajectories of the wound. The HealingIndex AI Model uses deep learning techniques to examine patient wound records, images, and characteristics, producing a score from the AI model that indicates the level of risk for delayed healing and potential deterioration [28].

 While traditional care pathways rely on clinicians identifying and escalating wounds that are deteriorating or at risk of deteriorating, the HealingIndex automates these processes using a predictive algorithm. This allows the wound care team to rapidly identify and prioritize wounds that require specialist input. This identification reduces the risk of deterioration and complications that necessitate hospitalization or amputation, reducing overall health care costs associated with wound complications [1,31,38]. Furthermore, the use of this AI tool can identify wounds at risk for slow healing, which may require an adapted wound management protocol. By reducing complications and adapting the treatment of slow-healing wounds, the AI tool can reduce the overall costs associated with wound care [31]. Early identification of wound deterioration is crucial as it allows timely intervention, preventing progression and potentially leading to significant savings for the health care system through avoidance of unnecessary costs [34]. 

The use of the predictive algorithm may also augment clinical decision-making by identifying wounds that are healing effectively and therefore may not need daily or thrice-weekly dressing changes. Reducing the frequency of dressing changes reduces supply costs and frees patient and provider time. Reducing the frequency of dressing changes to once a week, where clinically appropriate, can free up thousands of hours of nursing time over the course of a year [54]. Overall, the AI-powered DWCT can enhance clinical efficiency by automating escalation and review processes.  

Additionally, the DWCT platform provides opportunities for asynchronous collaboration. Specialists are able to review patient records, including high-resolution time-lapse photos, and provide recommendations without requiring geographic co-location or congruent timing. This enables the specialist wound care nursing team to optimize their time by reducing travel between sites, reducing administrative burden, and creating care plans that can be executed by staff located near the patient. This offers the benefits of not only efficiency but also equity, as access to best practice care and consultation is no longer limited to the fortunate few who can be reviewed by the specialist nurses in person. In rural and remote parts of Canada, which, as noted above, are struggling to recruit and retain skilled providers, these tools enable more patients to access specialist consultation and review without necessitating travel.  

AI-powered DWCT can support increased efficiency by (1) reducing time spent on manual assessment and measurement of wounds; (2) automating escalation by flagging deteriorating wounds for review, which enables prioritization of at-risk patients for intervention; and (3) enabling collaboration between frontline staff and specialist wound care resources without requiring geographic colocation or time alignment.  

Universality and accessibility are embedded within Canada’s Health Act, and equity must be centered as a priority.  Inequities in the Canadian health system exist, especially related to disparities in health outcomes for rural and remote populations, and disproportionately affect First Nations, Inuit, and Métis communities who face burdens associated with travel and costs [44]. Rural health care policies are often guided by urban care models, which can exacerbate these inequities for rural communities [24].

Many rural, remote, and Indigenous communities experience inequitable access and poor outcomes, which may be attributable in part to resource shortages. Additionally, human biases, racism, colonial legacies, and historically unbalanced relationships between service users and providers also play a key role. Nyugen et al [44] explicitly state “Colonization and historical intergenerational traumas... have plagued the survivors and later generations with physical and mental trauma,” and “...many Health Care providers do not acknowledge the impact of colonization on the Indigenous community and disregard the social determinants of health as explanations for illness.” This, in turn, manifests as culturally inappropriate health services, which deter people from seeking health care. 

 In this context, the AI-enhanced DWCT can support more equitable care by enabling community care providers with existing, trusting relationships to provide culturally safe care.  This approach is aligned with mitigation strategies proposed by Indigenous health experts who highlight the importance of community partnerships and provision of culturally safe care [44,45]. 

Furthermore, patients with darker skin tones disproportionately experience poor wound outcomes, which may be attributable to both inequities in social determinants of health and limitations in provider ability to accurately identify clinically relevant changes in tissue in pigmented skin [55].  Johnson et al [55] note that erythema is an excellent example, as its presentation varies based on skin tone, and historically, medical education has failed to account for diverse skin tones in training and published literature [21]. Darker skin tone is associated with overestimation of wound size when using standard “width X length” measurements [46]. In comparison, the DWCT with AI-augmented AutoTrace, AutoDepth, and SmartTissue is able to provide accurate measurements and can enable color calibration to adjust for light conditions and skin pigmentation [46]. Thus, the use of the AI-enhanced DWTC can support more accurate measurements and potentially overcome some of the implicit biases that health care providers manifest. 

While AI-enhanced tools offer immense promise to support better care efficiency and outcomes, there are several key considerations that must be addressed to ensure equitable and trustworthy implementation. One of the pressing concerns is the representation of diverse patient populations in AI systems. When datasets fail to capture diversity across factors such as skin tone, age, and other demographic variables, resulting models can produce biased or inaccurate outputs. These biases are particularly problematic in clinical contexts like wound care, where diagnostic accuracy is critical. Diverse groups, such as individuals with darker skin tones or older adults, are often underrepresented in training datasets, leading to biases in AI outcomes. This underrepresentation can result in inaccurate diagnoses or recommendations, which are harmful and counter to the goals of quality wound care. Human biases embedded in AI systems can become enshrined or even amplified if not intentionally designed with equity in mind [25]. Such biases arise when AI models are trained on datasets that do not adequately represent the diversity of the population, including those most likely to experience wounds, such as older adults [56]. Chu et al [56] emphasize the known challenges of ageism in AI, noting that models often overlook older adults due to insufficient data representation. To counteract these biases, it is essential to increase the diversity of AI training datasets and maintain robust evaluation protocols that ensure accuracy and interrater reliability across patient populations. These safeguards are crucial in mitigating the risks of bias and ensuring that AI systems are fair and inclusive. 

In addition to addressing bias, overcoming skepticism and implementation barriers is crucial for the successful adoption of AI-enhanced technologies, particularly among older adults. While the stereotype of older adults as technophobic is increasingly being disproven [57], structural barriers such as the cost of technology and access to reliable internet remain significant hurdles for many older adults in rural and remote communities [58]. Despite these challenges, patients are generally open to the use of AI in health care if appropriate regulatory and safety oversights are in place [59]. Pilot studies, such as those by Wang et al [52] and Raismam et al [60], show promising results regarding patient acceptance of AI for wound care [52,60] while Mohammed et al [34] highlight positive provider acceptance [34]. These studies indicate a growing receptiveness toward an AI-augmented DWCT, but continued exploration into unique barriers, especially those pertinent to the Canadian context, is necessary.  

The evaluation being conducted by this team will enhance understanding of implementing AI-enhanced DWCT in rural and remote communities in Canada. This evaluation will identify specific barriers and facilitators of adoption, ultimately leading to the creation of a framework that supports wider adoption and integration of technologies to support healthy aging. By addressing these issues, the path toward equitable, efficient, and widespread use of AI technologies in health care becomes clearer, fostering a more inclusive technological future.