Paid for. But unable to dispense.

Patients were ordering medication through uMeds, only to discover that the pharmacy could not dispense it. These orders were marked Unable to Dispense (UTD) and referred to Pharmacy Operations for investigation.

What began with a pharmacy visit in July 2024 became a two-year effort to understand the problem, advocate for prioritisation, and address its underlying causes. The solution combined an order allocation fix with a pharmacist-facing interval time visualiser.

Following the visualiser’s launch, weekly Interval Time UTDs fell from 97 to 52, with the improvement sustained over the next 13 weeks.

A successful checkout did not guarantee a dispensable order.

During a pharmacy visit in 2024, I observed pharmacies receiving orders they could not dispense. Each UTD created work across multiple teams. Pharmacy Operations investigated operational issues, Nurses resolved clinical issues, and Customer Support and CRM communicated the delay to patients. Cancellations also required Finance to process refunds.

As the backlog grew, patients could wait days for an update. Dispense Limit and Interval Time were the two largest recorded UTD categories. I repeatedly advocated for these issues to be prioritised.

What is interval time?

Interval time specifies how frequently a prescription repeat can be supplied. Consider a simplified prescription allowing two units per repeat, with a 14-day interval. Ordering the next repeat before the interval has elapsed can result in a dispensing restriction.

However, uMeds also supports partial quantities within a repeat: a patient might purchase one unit on Day 1 and the remaining unit on Day 7. Medicinal cannabis can cost around $100 per unit and is not subsidised through the PBS, making this flexibility important.

Internally, the second purchase is called an Interval Period in Progress (IPIP) order. Distinguishing an IPIP order from a genuine interval time order became central to the investigation.

The pharmacist could see the previous dispense, but not the whole context.

In April 2026, I interviewed Pharmacy Operations and engineers to investigate why Interval Time UTDs persisted. Two problems emerged.

An allocation rule broken by Click & Collect

Approximately 90% of partner pharmacies accepted IPIP orders, while 10% did not. Existing allocation logic accounted for this distinction. The Click & Collect launch introduced a gap allowing IPIP orders to reach pharmacies that did not accept them. An interface change alone could not resolve this source of preventable UTDs.

Missing context during dispensing

Pharmacists checked Real Time Prescription Monitoring (RTPM) as part of dispensing. They could see that the same medication had recently been dispensed, but not clearly establish that the new order represented the remaining quantity from the same repeat.

Rather than investigate the patient’s uMeds order history themselves, pharmacists referred these orders to Pharmacy Operations. The team then reconstructed the prescription and dispensing timeline and sent supporting evidence back to the pharmacy.

From a patient-facing intervention to a pharmacist-facing solution.

Fix the routing, then support the dispensing decision.

Restore IPIP-aware allocation

I worked with Delivery and Engineering to prioritise the Click & Collect allocation fix, preventing IPIP orders from reaching pharmacies that did not accept them. This addressed a system-level problem before asking pharmacists to change their behaviour.

Make interval time context visible

I designed and tested an interval time visualiser for pharmacists, then handed it over to Delivery and Engineering. It brought the relevant dispensing timeline into the pharmacist’s workflow, helping them assess whether an order represented a partial supply within an existing repeat.

The aim was to reduce the need for Pharmacy Operations to reconstruct and explain the same information manually.

“Easy to understand, the dispensed dates are accurate when cross-checked with RTPM.”

— Pharmacist feedback

Reducing the problem did not require eliminating every exception.

Some Interval Time UTDs remained because consecutive orders could be placed before the previous order was dispensed, before the next interval could be calculated. Pharmacists sometimes selected Interval Time when another reason applied. Pharmacies also differed on calculation when interval time changed between scripts.

By September the recorded rate was approximately 0.25%. With the issue more manageable, I prioritised the larger Dispense Limit UTD problem rather than pursuing increasingly marginal reductions.

The right intervention was not another checkout warning.

The first intervention helped patients understand when their order would be held. It did not resolve the context pharmacists needed to assess partial quantity orders.