Direct-to-patient Teledermatology and Computer-assisted Diagnosis

Overview

This study will examine access to dermatology care. The operational partner will implement a direct-to-new patient teledermatology clinical care pathway while the research will study the effects of that implementation, use data associated with deployment to develop and test an Artificial Intelligence system, and understand key stakeholders' attitudes regarding and readiness for remote dermatology care.

VA is uniquely positioned to develop and study direct-to-patient teledermatology in parallel with computer vision for skin disease. The following aims will assess the performance of these emerging patient-facing dermatology innovations and assess their ability to improve access to quality skin care VA-wide. 1. Assess the impact of direct-to-patient teledermatology on access and health system utilization. In collaboration with the VA operational partner, Office of Connected Care, the My VA Images app will be deployed to three VA facilities where the app will be introduced as an option to refer eligible new patients for dermatology consultation. Multiple access metrics, including time to consult completion and geographic distance traveled. will be measured in exposed patients and compared with patients referred to usual in-person and consultative teledermatology pathways. Facility-centric measures of access such as clinic appointment wait times and in-person dermatology clinic and community care utilization will also be measured. To better understand end-user experiences, the investigators will survey patients and staff at each study site to evaluate their satisfaction with My VA Images as well as the overall process. 2. Assess, refine and augment computer-assisted evaluation of patient-submitted images. An artificial intelligence-powered computer vision model, trained and validated on clinic-captured images of melanoma and nevi, will be tested and refined on patient-submitted teledermatology images from the My VA images app. The investigators will also extend the computer vision model by using all patient submissions to train and validate the model on a wider variety of skin diagnoses. Patient-submitted teledermatology images at study sites will be prospectively evaluated by the investigators' own computer vision model as well as by a commercial system currently available to VA clinicians and patients, and results will be compared with benchmark diagnoses to measure concordance across a range of diagnostic categories. 3. Assess readiness of VA and Veterans' acceptance to implement direct-to-patient care. The investigators will survey Veterans and key VA leadership and staff at three selected VA facilities, supplemented by interviews of key VA stakeholders, to understand patients' and organizational readiness, including facilitators and barriers, for transitioning to patient-facing technologies in general, and direct-to-patient dermatologic care in particular. National Veteran surveys, and VA Mobile Health user satisfaction data will help place local observations in perspective. Implementation and sustainability of the patient-facing teledermatology app technology will also be measured.

Conditions

Interventions

Eligibility

Locations (3)

Outcomes

Central Contacts