Background: Improving access to care is a high priority within the VA. While improvements to access have been made in recent years, gaps and inefficiencies still exist, particularly around missed clinic visits, or `no- shows'. The VA reports that approximately 15-18% of scheduled outpatient primary care appointments are not completed and that 9.2 million appointments were lost because of no-shows in FY2017. In preliminary work, we demonstrated the importance of social risk factors on VA no-show rates. These findings suggest that a no- show prediction model that incorporates patient-level factors could predict missed clinic rates and provide clinical phenotypes (i.e. an aggregate description of a Veterns' social vulnerabilities) of Veterans at greatest risk of no-showing. VA Video Connect (VVC) is a newly developed telemedicine application that provides video conferencing services as a means to connect Veterans with their VA medical providers. With VVC, Veterans can access their VA provider from any mobile or web-based device (e.g. smartphone, tablet, or computer) and do not need to be located at a satelite clinic. Previous work supports the idea that VVC could be targeted to those at elevated risk of no-showing clinic appointments. This CDA proposes a risk-based, targeted use of VVC in patients with social vulnerabilities as a means of decreasing clinic no-shows. Significance: This proposal aims to improve access to care by identifying, describing and engaging Veterans who would most benefit from alternative methods of primary care, specifically VA Video Connect. Innovation: This research has several innovative aspects to it. First, we will utilize machine-learning predictive techniques to identify and describe Veterans who are at highest risk of no-showing based on their social risk. This methodology has never been utilized in addressing no-shows. Second, we will actively engage Veterans in a formative assessment of how to optimize the use of VVC as an alternative method to obtaining primary care. Engaging Veterans throughout this proposal will ensure that Veterans' voices are properly integrated into the final product. Finally, this proposal utilizes novel telemedicine technologies (i.e. VVC) as a means of improving access for Veterans who are at high risk of missing clinic visits. Specific Aims & Methodology: (1) Use regression tree analysis to phenotype Veterans based on their estimated risk of no-showing clinic appointments. Hypothesis: Social risk factors are associated with no- shows in the ambulatory VA population and certain phenotypes will have higher no-show rates compared to others. (2) Use a sequential exploratory mixed methods design to engage phenotyped Veterans at high risk for no-showing and assess Veteran suitability and capability of using VVC. Hypothesis: Certain phenotypes of Veterans will be optimally served by VVC, while other phenotypes will require higher intensity primary care programs or continued in-person care. (3) Pilot the targeted use of VVC among 50 Veterans at-risk of no-showing primary care clinic appointments at the SFVA using a Type I hybrid effectiveness-implementation design. We will collect formative implementation data about local adaptability, acceptability, and fidelity. Hypothesis: VVC will be an acceptable alternative modality of primary care for both Veterans and providers. Next Steps: Following the effective implementation of this CDA, we will work with operational partners (Office of Connected Care and Telehealth) and perform a multisite assessment of the focused use of VVC on Veterans at high risk of missing clinic appointment.
NIH Reporter Project Information
None at this time.
Prevention, Technology Development and Assessment
None at this time.