PreventaTri Healthcare Technologies

Introduction

PreventaTri Healthcare Technologies delivers preventive, home-based care to medically vulnerable and Medicaid-eligible patients across Pennsylvania. Their model depends on frontline caregivers who visit patients in the community, and on a clinical team of registered and licensed nurses (RN/LPN), case managers, and administrators who supervise care across counties and regions.
The core clinical challenge in this model is time: catching a patient’s deterioration early, before a preventable condition becomes an emergency room visit or hospitalization. PreventaTri needed the technology to match that mission.

Challenges

Before partnering with Desol Int, PreventaTri’s care coordination ran on fragmented, largely manual processes that created real clinical risk:

• Unstructured observations. Caregiver check-ins were inconsistent and paper-based, with no standardized way to capture what was observed at the patient’s home — and no objective way to score how concerning it was.
• No early-warning signal. A deteriorating patient could go unnoticed until the next scheduled review. There was no automated mechanism to surface the highest-risk patients to the clinical team in real time.
• Broken accountability chains. When something urgent was flagged, there was no guaranteed path to a responsible clinician. Alerts fell through the cracks — no one owned the response, and no one could prove who was notified or when.
• Complex geographic coverage. Assigning the right nurse to the right patient across dozens of Pennsylvania counties and regions was manual and error-prone.
• Compliance and auditability pressure. As a Medicaid provider handling protected health information, PreventaTri needed rigorous access control, a defensible audit trail, and data isolation between organizations — not spreadsheets and email.

They needed a single platform that could turn everyday caregiver observations into an objective, real-time clinical safety net.

Solution

Desol Int designed and built a multi-tenant, role-based patient monitoring and escalation platform that closes the loop from a caregiver’s observation to a clinician’s response — automatically.

The platform is organized around four distinct roles, each with a purpose-built experience and strictly scoped data access:

• Caregivers complete structured assessment checklists during home visits and can raise immediate red-flag alerts — without ever seeing other patients’ clinical data.
• Nurses (RN/LPN) receive risk-ranked alerts for the patients and regions they cover.
• Case managers conduct supervisory visits and follow-up tasks within their assigned regions.
• Administrators configure assessment templates, manage staff and geography, and oversee the whole organization.

Admin can view system-wide statistics, including the total number of nurses, total number of caregivers, and the total number of red flags raised. Nurses, Case manager, Cargiver can also view statistics specific to their accounts, including the total number of patients assigned to them, the total number of red flags, and the total number of yellow flags. A notification system is implemented to ensure timely alerts. When a red flag is raised, the relevant person is notified. If the red flag is escalated, the next responsible person in the escalation hierarchy is also notified. Additionally, a follow-up task can be generated to ensure the patient’s condition is reviewed and appropriate action is taken.

An automated triage engine. Every submission is scored by a risk engine that assigns a GREEN / YELLOW / RED flag and a risk category. Instead of a clinician manually reviewing every visit, the system continuously surfaces the patients who need attention now — turning routine check-ins into a live triage board.

A TTL-based auto-escalation ladder. This is the platform’s clinical safety net. When a RED flag is raised, the system routes it through a defined chain of responsibility — patient’s primary nurse → regional primary nurse → secondary nurse → regional secondary nurse → case manager → administrator — with a time-to-acknowledge deadline at each level. If a responder doesn’t acknowledge in time, the alert automatically escalates to the next tier. Every hand-off is recorded, so a critical alert can never silently die, and the organization has a complete, provable record of who was paged and when.

Implementation Details

Desol Int delivered a modern, production-grade stack engineered for reliability, security, and scale:

Frontend

• Next.js 16 / React 19 with TypeScript for a fast, type-safe, server-rendered dashboard
• Material UI (MUI) component system with a consistent design language and data-dense grids
• TanStack React Query for efficient data fetching and caching.
• Socket.io client for real-time in-app notifications

Backend

• Node.js / Express REST API with a versioned, documented (Swagger) surface
• PostgreSQL with Prisma ORM as the system of record, modeling a rich clinical domain — patients, nurses, caregivers, case managers, regions/counties, checklist templates and submissions, flags, escalations, supervisory visits, and follow-up tasks
• Socket.io for real-time event delivery
• AWS S3 for secure storage of image and audio attachments

Security, Compliance & Reliability

• Multi-tenant data isolation — every record is scoped to its organization, so no data ever crosses tenant boundaries
• Granular role-based access control and per-user permissions
• OTP-based two-factor authentication, account lockout on repeated failed logins, and sliding, idle-aware session timeouts
• Comprehensive audit logging of every sensitive action (create, read, update, delete, login, password reset, account lock/unlock) — capturing actor, before/after state, and IP
• Idempotency controls to guarantee that retried requests never double-process a critical clinical action
• Soft-delete architecture preserving historical records for compliance and recoverability
• AWS RUM & X-ray for real-time error tracking and real-user performance monitoring

Because escalation timing is clinically important, particular engineering care went into the auto-escalation state machine — a TTL-driven engine that reliably advances alerts through the responsibility ladder and preserves a tamper-evident event history at every step.

Outcomes

The platform transformed PreventaTri’s care model from reactive and manual to proactive and data-driven:

• Earlier intervention. Structured, scored assessments mean at-risk patients are surfaced the moment a caregiver logs a concern — [reducing time-to-clinical-review from days to minutes].
• No dropped alerts. The auto-escalation ladder guarantees every RED flag reaches a responsible clinician, eliminating the accountability gaps that previously let urgent cases stall. [X%] of critical alerts are now acknowledged within the target window.
• Higher clinical productivity. By triaging automatically, nurses focus their time on the patients who need it most instead of reviewing every routine visit.
• Audit-ready compliance. A complete, immutable audit trail and strict multi-tenant data isolation give PreventaTri the governance posture required of a Medicaid provider handling protected health information.
• A scalable foundation. The multi-organization architecture and region/county coverage model let PreventaTri expand to new teams and geographies without re-platforming — supporting [N patients across N counties] on a single system.

Why It Matters

For a preventive-care provider, technology isn’t overhead — it’s the difference between catching a problem early and reacting to a crisis. Desol Int delivered PreventaTri a platform that turns thousands of everyday caregiver observations into a real-time clinical safety net, with the security and auditability their industry demands.

Interested in a similar solution? Desol Int builds secure, scalable healthcare and enterprise platforms end to end. [Contact us] to discuss your project.