Case Study: Patient Access Appointment + Request Assistant (Phone/SMS/Web) | Envision360
Case study

Patient access appointment + request assistant (Phone/SMS/Web)

Client: regional health network (three sites). A single intake layer that handles inbound calls and messages, books appointments, routes “non-booking” requests to the right department, and runs outbound reminders with clean staff handoff. Built to work in the U.S. and Canada, and to sit beside the EHR (including MyChart-style workflows where applicable).

Client at a glance

  • Operations: multi-site outpatient + hospital-affiliated departments with centralized scheduling
  • Channels: phone (voice), web chat, SMS; optional WhatsApp
  • Request types: scheduling/rescheduling, prep instructions, directions, billing questions, records requests, FAQs
  • Goal: reduce phone tag, shorten time-to-book, and stop requests from bouncing between teams
Patient access work queue and scheduling dashboard interface

Challenge

The EHR held the chart. The problem was everything around it: calls, transfers, and requests that weren’t owned end-to-end.

  • High call volume: long holds, abandoned calls, and repeat callers
  • Transfers: a simple question often bounced through multiple departments
  • Reschedules: gaps in calendars and last-minute cancellations left unused capacity
  • Inconsistent reminders: staff-dependent follow-ups meant more no-shows
  • Missing context: intake details arrived incomplete, causing back-and-forth
  • After-hours backlog: requests piled up overnight and hit staff in the morning
Mobile and web views of the patient access assistant flow
Our solution

One intake layer that books, routes, and closes the loop.

Booking and rescheduling

Patients can book or move appointments by phone, chat, or SMS. The flow collects the right details, then checks live availability with rules to prevent bad bookings.

  • Visit type + location rules (provider, clinic, room, equipment)
  • One-tap reschedule/cancel with immediate calendar updates
  • Waitlist fill when a slot opens

Request capture + routing

Not every call is “make an appointment.” This layer captures common hospital requests and routes them to the right queue with a clear summary.

  • Prep instructions, directions, clinic questions, billing, records requests
  • Rules for urgency, hours, and escalation
  • Status tracking: received → in progress → complete

Example: we built a hospital workflow assistant that handled patient FAQs and routed common requests into the correct departmental queues, with staff handoff notes and tracking.

Reminders and recovery

Outbound messages reduce no-shows and cut inbound “I forgot” calls. Patients can confirm, reschedule, or ask for a call-back.

  • Reminders at set intervals (ex: 48h / 24h / 2h) with confirm/modify
  • No-show recovery proposes new times automatically
  • Prep and day-of instructions for imaging/procedures

Handoffs and guardrails

When a human needs to step in, staff get a short summary and the captured details—so they don’t restart the conversation.

  • Clear “handoff to staff” triggers (clinical questions, complex billing, sensitive cases)
  • Logged actions and audit trail
  • Built to avoid giving medical advice
How it fit their stack

Works with existing systems. Keeps data tight.

Channels

  • Phone (voice) with safe handoff to staff
  • Web chat widget and responsive flows
  • SMS via Twilio; optional WhatsApp
  • Accessible UI patterns

Routing + workflow layer

  • Department queues and routing rules (hours, urgency, required fields)
  • Templates for request types so details are captured consistently
  • Retries, timeouts, and “safe fallback” paths
  • Staff console with role-based access and audits

Integrations

  • EHR patterns: FHIR/HL7 when available (or scoped connectors where practical)
  • Scheduling: Google/Microsoft 365 calendars when used
  • Notifications: SMS (Twilio), Email (SendGrid)
  • Payments: Stripe/Square (optional, visit-type dependent)

Security + privacy

  • PHI minimized in logs; sensitive fields handled carefully
  • Encryption in transit; access controls; audit trails
  • Data residency options (Canada or U.S.)
  • Designed to align with PHIPA/PIPEDA and HIPAA expectations

Hospital platforms we commonly design around: Epic / MyChart, Oracle Health (Cerner), MEDITECH, athenahealth, Allscripts/Veradigm, NextGen, eClinicalWorks (based on access and feasibility).

Pilot

Scope and constraints

  • Eight-week pilot across three sites (appointments + the top request categories driving calls)
  • Different departments, different rules: routing logic reflected real hours, escalation, and “must-have” fields
  • Identity checks: DOB/contact match or short verification code where needed
  • Edge cases: referrals, pre-auth questions, and clinical topics always routed to staff with a clean summary

Result: fewer repeats, fewer transfers, and a steadier schedule with less manual chasing.

Impact

First 60–90 days (pilot)

  • More completed bookingsHigher conversion vs phone-only baseline
  • Fewer abandoned callsPatients could self-serve or leave a structured request
  • Lower no-show rateConsistent reminders + easy rescheduling
  • Less front-desk timeFewer “restart the story” calls
  • After-hours coverageRequests captured overnight instead of piling up
  • Better visibilityQueues showed what was open and who owned it

Note: outcomes vary by department, patient mix, and how much is integrated vs handled as a tracked request.

Roadmap

What was next

  • Deeper write-backs where the EHR supports it (appointments, messages, request status)
  • More departments: imaging prep + confirmations, HIM records requests, billing callbacks
  • Multi-language flows and accessibility enhancements
  • Smarter routing based on request history and department capacity
  • Insurance/eligibility checks for select visit types (when access is available)

Client: name withheld — regional health network with three sites.

Schedule a call or contact us with one workflow you want to reduce.