Bringing pharmacy fulfillment to the bedside
I led the UX architecture for a mobile hospital pharmacy system that translated a readmission-reduction goal into a bedside fulfillment model grounded in pharmacy workflow.
6 transaction types · 5 pharmaceutical categories · up to 8 signatures
Patients were sometimes leaving the hospital without medications essential to their recovery. Patient Engagement Advisors wanted to increase prescription fulfillment before discharge and reduce avoidable readmissions — the opportunity was to bring pharmacy fulfillment directly to the patient's room, on a unified mobile experience.
I led the UX architecture, directed the research, and co-designed the interaction and final visual direction. One continuous full-time UX lead, supported part time by research, design, product, and implementation partners, across a nine-week engagement. This case traces how that outcome — more patients leaving with the medications they needed — became product requirements, an interaction architecture, and a final design direction.
The Challenge
The prior system could process a transaction, but it could not reliably support bedside fulfillment. Patient context, transaction state, and payment each lived in their own modal. Closing a modal repeatedly obscured the patient's status and forced technicians to re-establish context. The workflow was also fragmented across digital and analog systems: some signatures were captured within specific transactions, like credit-card payment, while many pharmaceutical and compliance acknowledgements remained on separate paper forms stored behind the pharmacy counter. Technicians had to remember which requirement applied, retrieve the correct form, and reconnect that documentation to the active patient and transaction. The current product architecture could not reliably support the service model the hospital wanted to deliver: bedside fulfillment meant bringing that paper-based compliance process into the same digital transaction technicians were already juggling across modals.
Before: one transaction split across two systems
On screen: context obscured by modal workflows
Patient context hidden
Product and transaction state obscured
Origin, progress, and next action unclear
Off screen: acknowledgements managed on paper
Privacy acknowledgement
Pharmacist consultation
Safety-cap preference
Generic illustrative reconstruction, not an original hospital document
Privacy, consultation, and pharmaceutical acknowledgements were managed on separate paper forms stored behind the pharmacy counter. Technicians had to determine which forms applied, retrieve them, capture signatures, and reconnect the documentation to the active patient and transaction.
Research & Reframing
I directed ethnographic research across two hospital pharmacies, observing how technicians actually worked the floor — the paperwork, compliance and signature requirements, payment variability, and how pharmaceutical category changed what a transaction required. That variation wasn't an edge case; it was the baseline of every shift. It reframed the goal into a single patient-transition workflow — the operational sequence technicians needed to complete as one continuous bedside transaction:
patient
needs
exceptions
assistance
confirm
discharge
Six stages of one patient-transition workflow — not six separate tools.
Translating the outcome into a product model
The hospital's objective wasn't simply to deploy a mobile POS — it was to increase the number of patients who left with the medications they needed. I translated the hospital's outcome into specific product conditions the interface had to support:
Those requirements became the interaction architecture.
Interaction Architecture
The spatial model replaced modals with fixed, predictable regions. Patient, prescription, basket, and transaction state stayed visible in a persistent center workspace, designed to reduce the need to reconstruct status mid-transaction. Contextual controls surfaced from the left based on the active task.
Signature and confirmation entered from the right, consolidating requirements previously split between specific digital transactions and paper forms behind the counter into one interaction layer, so compliance became part of the active transaction rather than a separate paper process. Critical alerts entered from the bottom, so blocking issues were visually distinct from requirements that could wait.
A spatial interaction model for preserving context
Global controls
Temporary workspace
Contextual controls
Modifies the active input workspace
Primary input workspace
Persistent base layer
Basket and payment
Persistent base layer
Signature and confirmation
Extends the active basket and payment state
Critical system alerts
Interrupts and locks the active workspace
One architecture, reused across every transaction type, rather than a custom flow for each.
Final Design
From interaction model to final design
The final design translated the spatial model into a stable workspace that preserved patient and transaction context while temporary actions appeared only when needed.
Final design direction
-
Patient and input workspace
Patient and prescription context remained visible in the design throughout the transaction.
-
Persistent basket and payment
Transaction contents and payment state were designed to stay available throughout related tasks.
-
Medication status and exceptions
Requirements, exceptions, and blockers were surfaced within the active workflow instead of being hidden behind separate screens.
-
Contextual patient actions
Secondary profile actions — editing details, notes, and purchase history — were designed to stay reachable without leaving the active transaction.
Reported Outcomes
Project reporting from the first hospital implementation showed orthopedic-surgery readmissions declining from 7% to 4.4%, alongside increased prescription capture.
The hospital's underlying methodology and longer-term data were not available to me, so this is presented as a reported implementation outcome rather than a causal attribution to design alone.
This project demonstrates how I translate an organizational outcome into the workflow, information architecture, and product behavior required to support it—while preserving the operational context users need to complete complex healthcare transactions.