How One Healthcare Company Found Its Way With a PoC

This article outlines a composite case study, drawing from common challenges and successful strategies we’ve observed across multiple healthcare engagements. It’s not a single client story, but a blend of patterns. Our hypothetical client, "MediCare Innovations," was a regional healthcare provider network based in the Great Lakes region, managing 15 clinics and two community hospitals, specializing in integrated primary and specialty care.

The Starting Point

MediCare Innovations faced a growing problem: patient no-shows and late cancellations for specialty appointments. These missed appointments didn't just impact revenue; they delayed care for other patients, wasted physician time, and created operational bottlenecks for scheduling staff. Their existing system, a decades-old Electronic Health Record (EHR) from Epic with a custom-built patient portal, offered basic appointment reminders via email and SMS. However, these were generic, one-way notifications, easily overlooked amidst a barrage of digital communication. Their patient engagement team reported that 37% of new specialty appointments were either missed or canceled within 24 hours of the appointment, costing the network approximately $1.2 million annually in lost revenue and administrative overhead. Scheduling staff spent an average of 15 hours per week manually calling patients to confirm appointments or reschedule. They recognized the need for a more dynamic, personalized reminder system but were hesitant to commit to a full-scale development project without proving its efficacy and user acceptance.

What They Shipped

To address the no-show problem, MediCare Innovations partnered with us for a Proof of Concept (PoC). The goal was to build a focused, functional prototype that could be tested with a small group of real patients and staff, validating the concept before a larger investment. Over 10 weeks, we delivered:

1. Two-way SMS Confirmation Module: A prototype feature integrated with their existing patient data via a secure API gateway, allowing patients to confirm, reschedule, or cancel appointments directly through a simple text message. This module used Twilio for SMS delivery and Amazon Web Services (AWS) Lambda for backend logic.
2. Personalized Reminder Logic: The system dynamically adjusted reminder frequency and content based on appointment type (e.g., surgical consults received reminders 7 days, 3 days, and 24 hours prior; routine follow-ups received 3 days and 24 hours prior) and patient history (e.g., patients with a history of no-shows received an additional reminder).
3. Simple Staff Dashboard: A web-based interface built with React, providing scheduling staff a real-time view of appointment statuses (confirmed, pending, canceled, rescheduled) and a quick way to manually intervene or send custom messages if needed. This dashboard connected to the same backend logic as the SMS module.
4. Integration Mock-up for EHR: While not fully integrated, we created a clear architectural diagram and a set of API specifications demonstrating how the PoC’s core functionalities would eventually connect with their Epic EHR’s scheduling module, specifically outlining data flows for patient demographics and appointment slots.
5. User Feedback Collection Mechanism: Embedded within the prototype was a simple, anonymous survey tool for both patients (via a link in the confirmation message) and staff (within the dashboard) to gather immediate usability and satisfaction data.

The Numbers, 6 Months In

The PoC was rolled out to a pilot group of 500 patients scheduled for dermatology and cardiology appointments over three months. Following successful initial feedback, it was expanded to cover 2,000 patients across four specialty departments. Six months after the initial pilot launch, MediCare Innovations observed significant improvements:

• 18% Reduction in No-Shows: The overall no-show rate for specialty appointments dropped from 37% to 19% within the pilot groups.
• 30% Decrease in Manual Scheduling Calls: Scheduling staff reported spending 30% less time on manual confirmation calls, freeing up approximately 4.5 hours per week for other patient support tasks.
• 92% Patient Satisfaction with Reminder System: Anonymous patient feedback indicated a high level of satisfaction with the new two-way communication, citing convenience and clarity.
• Projected Annual Savings of $450,000: Based on the reduction in no-shows and administrative burden, the network projected significant annual savings, recouping the PoC investment within the first year of a full-scale rollout.

What We'd Do Differently

Looking back, there are two areas where we could have refined our approach:

1. Earlier Focus on Edge Cases: While the core functionality was robust, we spent a bit too long on the "happy path" and not quite enough time on less common scenarios during the PoC phase. For instance, handling multiple appointments for the same patient on the same day, or specific consent requirements for certain types of care, became more complex than anticipated during the pilot. Building out a few more edge-case flows earlier might have streamlined the subsequent full development planning.
2. More Granular A/B Testing: The PoC validated the overall concept, but we could have incorporated more robust A/B testing within the prototype itself to fine-tune reminder timings or message phrasing. For example, testing "Your appointment is tomorrow at 2 PM. Reply Y to confirm" versus "Quick reminder: Dr. Smith tomorrow at 2 PM. Confirm?" might have yielded even better engagement rates before scaling. This would have required slightly more upfront design complexity but offered richer data.

Other healthcare teams grappling with similar operational inefficiencies could benefit from MediCare Innovations' approach: start small, focus on a single, measurable problem, and use a functional prototype to gather real user feedback. This iterative validation minimizes risk and provides concrete data to justify larger investments in care-grade software.