A HIPAA-compliant mobile care platform for underserved populations — translating NIH-funded research into real-world clinical outcomes by solving for access, engagement, and behavior simultaneously.
Despite nearly all people living with HIV being aware of their diagnosis, the U.S. HIV care continuum revealed a persistent, systemic breakdown: the further patients moved along the continuum toward viral suppression, the steeper the drop-off.
The gap wasn't awareness — it was sustained engagement. Patients were falling out of care due to geography, poverty, stigma, distrust, and the practical impossibility of showing up at a clinic when life is unstable.
Existing tools — SMS texting, patient portals, phone calls — were insufficient. They were one-directional, literacy-dependent, or simply not used. A new model was needed: one designed for the realities of this population, not the ideals of a clinical trial.
Prevalence-based HIV Care Continuum
U.S. and 6 Dependent Areas, 2019
Percent of all people living with HIV
Note: Receipt of care = ≥1 test (CD4 or VL) in 2019. Retained in care = ≥2 tests ≥3 months apart. Viral suppression = <200 copies/mL on most recent test.
of persons with diagnosed HIV infection remain retained in ongoing care — the critical gap CAREiSCOPE was designed to close.
The problem wasn't awareness — 87% of PLWH knew their diagnosis. The gap was sustained engagement with care after diagnosis.
CAREiSCOPE grew from a decade of NIH-funded mHealth research at the University of Virginia's Infectious Disease Clinic (Ryan White Part C, 750 PLWH). I joined as product lead in partnership with Dr. Rebecca Dillingham and Dr. Karen Ingersoll, translating their formative research into a deployable clinical product.
The PositiveLinks research program — two formative pilots from 2013–2017 — provided the evidentiary foundation. Targeted at patients most at risk: newly diagnosed, history of missed visits, below the federal poverty line. Medically complex, multiple comorbidities, 67% with behavioral health diagnoses.
This wasn't a consumer app looking for a clinical use case. It was a clinical system built from the ground up to answer one question: what does it actually take to keep the highest-risk patients in care?
Both pilots targeted highest-risk patients. Pilot outcomes published across four peer-reviewed papers
Co-developed with UVA School of Medicine faculty including Dr. Dillingham (ID), Dr. Ingersoll (Psychiatry), and Dr. Flickinger (General Medicine) — each contributing domain expertise to the behavioral and clinical design.
From day one, the design principle was that the platform had to work for the most marginalised patients — not the median patient. That constraint produced a more robust, more equitable product.
CAREiSCOPE was designed as a three-layer system: a patient-facing iOS/Android app, a HIPAA-compliant cloud infrastructure, and a web-based clinical administrative portal. Every component was purpose-built to work in concert — not assembled from off-the-shelf tools.
The design philosophy was rooted in three principles borrowed directly from the research: language should not be a barrier, location should not be a barrier to care, and literacy should not be a barrier.
This meant real-time translation to 100+ languages, telemedicine-capable video visits, HIPAA-compliant document handling, and an interface designed for low digital literacy — all without sacrificing security or clinical utility.
The most consequential design decision in CAREiSCOPE wasn't a feature — it was a behavioral economics model that reframed the problem of mobile data access as a reinforcing incentive loop.
For 72% of CAREiSCOPE's target population, access to mobile data was not guaranteed. Standard digital health products treat this as a deployment limitation. We treated it as a design constraint to solve.
Patients complete daily check-ins, medication queries, and community interactions — generating engagement data for the clinical team.
Engagement is rewarded through the optional RealHealthRewards program: $50/month in mobile data reimbursement when response rate exceeds 50%.
Mobile data access enables patients to reach health resources, messaging, appointment coordination, and peer support — between clinic visits.
Sustained engagement drives viral suppression, retention in care, and reduced avoidable utilization — reinforcing the loop with measurable health benefit.
The key insight: mobile data wasn't just a distribution channel — it was a social determinant we could directly address. By making health engagement the mechanism for data access, we created a self-reinforcing loop in which the act of staying in care funded the ability to stay in care.
Outcomes from two PositiveLinks pilots were published across four peer-reviewed papers. The results demonstrated that a well-designed mobile platform could move clinical outcomes in a population that the healthcare system had failed to retain.
CAREiSCOPE was designed from the infrastructure layer up for clinical deployment — not retrofitted for HIPAA compliance after the fact. Every system component was architected around the assumption that it would handle protected health information for the most vulnerable patient populations.
The three-tier architecture — member app, HIPAA-compliant secure cloud, and clinical administrative console — allowed bidirectional data flow between patients and care teams, while maintaining strict access controls and audit logging throughout.
The admin console exposed an API layer for EMR and HIT integration, enabling CAREiSCOPE to function as a connective tissue layer between patient-facing engagement and institutional clinical systems — rather than a standalone silo.
Had I had this app when I was newly diagnosed, I believe wholeheartedly that I would not have struggled as much. It is very important to know that someone cares. I can be in constant contact with my whole team.
The decisions we made for the most vulnerable patients — low literacy, no reliable data, high stigma — produced a more robust, more usable product for everyone. Designing for the median user produces median outcomes.
The incentive loop wasn't a marketing feature — it was a systems design decision that solved a social determinants problem. When behavioral theory is embedded in the product architecture, not layered on top of it, the outcomes follow.
Most digital health products lack clinical evidence for what they claim to do. Having peer-reviewed outcomes data changed every conversation — with funders, with health systems, with regulators. The research investment was a product asset.