Case Study - Smart Prescribing

Smart Prescribing

Redesigned digital prescriptions by observing real doctor workflows—reducing typing, accelerating consultations, and improving feature adoption.

Background

Ayukoash was an early-stage healthcare technology startup focused on creating comprehensive digital health profiles for patients. The vision was to digitize and streamline clinical workflows while improving continuity of care, data accuracy, and long-term health outcomes.

One of the most critical components of the platform was digital prescription creation, a feature intended to replace handwritten prescriptions and serve as a structured, reusable medical record. The application was primarily used by doctors on tablet devices during live patient consultations.

Despite positive feedback on the overall platform, adoption of the digital prescription feature remained inconsistent. Some doctors continued to rely on handwritten prescriptions, later uploading them or bypassing the feature entirely.

Given that prescriptions were central to both patient records and downstream workflows, low adoption posed a strategic risk to the product’s long-term value.

Problem Statement

Core Challenge

How can a digital prescription system be designed to match—or exceed—the speed, flexibility, and familiarity of handwritten prescriptions, without disrupting doctors’ existing consultation workflows?


At a surface level, the problem appeared to be one of usability. However, deeper investigation suggested that the issue was not merely interface design, but a misalignment between digital systems and entrenched clinical behaviors.

Initial Assumptions and Constraints

Assumptions
  1. Doctors would be willing to type prescriptions if the interface was “simple enough”
  2. Standardized digital forms would improve clarity and reduce errors
  3. Structured data entry was inherently superior to free-form writing
Constraints
  1. Doctors had limited time per consultation
  2. Typing on tablets was slower than writing
  3. Resistance to behavioral change in clinical settings is high
  4. Accuracy and reliability were non-negotiable due to medical implications

Discovery and Research Approach

Methods Used
  • Direct observation of doctors during real consultations
  • Shadowing prescription-writing behavior
  • Informal, open-ended conversations with doctors
  • Analysis of past prescriptions for patterns

This approach shifted the focus from what doctors said they wanted to how they actually worked.

Key Insights

Prescribing is Pattern-Based, Not Free-Form

Doctors rarely create prescriptions from scratch. They rely on familiar drug combinations and repeated templates for common conditions.

Follow-Up Visits Depend on Historical Context

Doctors frequently refer to previous prescriptions, modifying dosages or changing one or two medicines.

Instructions Are Highly Repetitive

Dosage and usage instructions follow predictable patterns and change infrequently.

Typing Was the Core Bottleneck

Manual text entry, repetition, and context switching disrupted clinical thinking.

Strategic Reframing of the Problem

Instead of asking “How do we make doctors type faster?”, the team reframed the problem to designing a system that anticipates intent, minimizes effort, and preserves clinical flow.


Outcomes

Internal testing and pilot usage demonstrated reduced prescription creation time, improved comfort among doctors, increased feature adoption during live consultations, and positive feedback on the system “feeling familiar” rather than “feeling digital.”

The feature became a proof point for the startup’s ability to design clinically empathetic technology.