Patient Communication Apps for Diagnostic Imaging Networks

Summary

We build patient communication applications for diagnostic imaging networks. A system of this kind gives the patient one place to call or message the referring practice, complete a referral form, and read or watch preparation material specific to the modality they are booked for. For the network it also standardises the referral intake path, which is otherwise spread across fax, phone and unstructured email, and it gives staff a single surface for publishing and retiring patient-facing content.

The Challenge

Imaging networks lose capacity whenever a scan is abandoned partway through. The common causes are claustrophobia in a closed bore magnet, pain that prevents a patient holding a position, and motion that renders a sequence unusable. Most are addressable before the appointment through preparation and contact, so the hard part is reaching the patient, not the imaging. Anything identifying a patient alongside a clinical context is protected health information, so HIPAA applies end to end. Referrals vary by payer context and resist fixed columns, and forms fail when validation is too strict for real clinical data. Notification permissions are declined often enough that a push-only design strands patients, and imaging centres frequently have poor signal, so a submission that is not queued is lost.

The Solution

Service API and clients

The usual shape is a service API plus a cross-platform client. The API is commonly built on .NET and C# over SQL Server, hosted in AWS, exposing endpoints for referral submission, content delivery, notification dispatch and admin content management. The mobile client historically used Xamarin Forms for a single shared codebase across Android and iOS; the same architecture now targets .NET MAUI, and the platform-specific parts are the ones that matter: telephony intents, notification registration and file handling. Alongside the app there is a responsive web surface, because referrers work from a desktop and will not install anything. An admin portal lets staff publish educational documents, associate them with a modality and a body region, and retire outdated versions without a release.

Referral model and integrations

The referral is the central entity, and it varies by payer context: group health, workers compensation, self-insured and self-pay referrals capture different fields and route to different queues. The data model treats the form as a versioned template with typed fields, so a new payer type is configuration. Communication channels are delegated:

• Click-to-call hands off to the device dialler.
• SMS goes through a messaging gateway and email through a transactional provider.
• Each channel records a delivery event against the referral, so staff can see what was attempted.

Scheduling and appointment context typically arrive from a RIS, and where imaging results are involved, HL7 v2 messaging or FHIR resources such as ServiceRequest, Appointment and DocumentReference are the integration surface.

Compliance and security

HIPAA drives encryption in transit and at rest, role scoped access, immutable audit logging of every read of a patient record, signed business associate agreements with every cloud and messaging vendor, and automatic session expiry on the mobile client. Push notification payloads never contain clinical detail, since they render on a locked screen. Educational content itself is not protected health information, so that store stays logically separate and cacheable at the edge.

How we build it

We separate the API and the mobile client into distinct deliverables with a contract between them, because app store submission latency should never block a backend fix. Referral templates and content are modelled as data from the start. Accessibility is treated as a requirement rather than a pass at the end, since many imaging patients are older or in pain when using the app. Forms support partial saves and resume, email and SMS act as fallbacks when notifications are declined, content is versioned so a patient can be shown the preparation sheet they actually read, and submissions queue rather than fail when signal drops.

What This Delivers

Patients arrive prepared for the study they are booked for, with a direct line to the practice when something is unclear. Referrers get one structured intake path instead of fax and free-text email. Staff publish and retire preparation material themselves, and can see which contact attempts were made against a referral. The network gains an auditable record of patient-data access that stands up to a HIPAA review.

Technologies and Tools

.NET and C# services over SQL Server hosted in AWS; Xamarin Forms and .NET MAUI for the shared Android and iOS client; a transactional email provider and an SMS gateway; RIS integration over HL7 v2 or FHIR.