Background: Access to expert skin care remains a challenging goal in the Department of Veterans Affairs (VA). While VA has successfully used consultative teledermatology between primary care providers (PCPs) and dermatologists to improve access to dermatologists, the reliance on PCPs is limiting and Veterans must still wait for and take time to travel to imaging appointments. An innovative path for VA is to develop and validate approaches to direct-to-patient dermatologic care. First, VA can augment its MyVAImages mobile app to allow Veterans to seek teledermatology consultation directly from VA dermatologists without the need to go through a PCP or a non-VA dermatologist. Second, VA can validate artificial intelligence (AI) methods for automatically evaluating skin diseases specifically in the Veteran population.
Significance/potential impact: The envisioned research will establish the operational and scholarly foundations to justify and enable VA to move toward a vision where Veterans have access to rapid expert dermatologic care wherever and whenever they need it. The proposed planning and research will study the impact of a growing trend where, instead of patients traveling to fixed clinic locations for healthcare, healthcare is delivered to them. While dermatologic evaluation is the focus of this research, the results will be relevant to other types of telehealth and AI approaches. This proposal will address the priority area, Access to Quality Care.
Specific goals for Phase I (Planning Phase): Aim 1 is to assemble the methods and instruments for assessing preferences and attitudes of both patients and VA staff toward direct-to-patient care. These will include development of methods for identifying and reaching subjects, surveys and interview protocols, and storing data securely. Aim 2 is to design a controlled trial for measuring the impact of allowing new patients to be directly referred for teledermatology evaluation on patient access to skin care. Activities will include ensuring access to pertinent data and collaborating with the operational partner, Office of Connected Care, to establish clinical workflows and support. Aim 3 is to establish an image data pipeline from the mobile app image repositories to our academic artificial intelligence computing environment, annotate previously archived images to further train our artificial intelligence model, and establish criteria for measuring concordance between AI and teledermatologists.
Long-term goals for Phase II (Research Phase): Aim 1: Assess the readiness of VA and Veterans for direct-to-patient care via teledermatology and AI image analysis. Aim 2: Measure the validity and impact of patient self-referral via teledermatology on access outcomes. Aim 3: Validate AI-mediated computer vision of patient-submitted teledermatology images and measure concordance with teledermatologists.
Innovation: 1) While direct-to-patient telemedicine and machine learning tools are not new in the field of health care, the proposed research will make a novel contribution in understanding how their implementation may be affected by all stakeholders’ readiness for novel and remote forms of health delivery. 2) Direct-to-patient teledermatology is completely unprecedented in VA and has not been well-studied outside VA. Using comprehensive system-wide data that is not typically available to most organizations, the proposed research will measure the technology’s impact on multiple dimensions of healthcare access. 3) While teledermatology and AI will not necessarily be used together when fully mature, their combined study in this project will allow an unprecedented opportunity to compare two technologies that will likely evolve together initially and eventually compete with each other. 4) At its conclusion, the research may validate two approaches for augmenting dermatologic care that may potentially be operationalized by Office of Connected Care.
External Links for this Project
Grant Number: I01HX002450-01
None at this time.
TRL - Applied/Translational
None at this time.