Retiring a twelve-step workaround, one practitioners had already built themselves
1 · Context
I ran the user research (interview protocol, practitioner interviews, survey analysis, and reporting UX findings back to the team), produced and presented the designs, built an interactive prototype, and designed the auto-generated compliance justification and its related alerts.
A new agreement introduced an obligation for physiotherapists: the nature of a prescription renewal now had to be recorded on the invoice sent to National Health Insurance, using a fixed classification (renewal identical or adapted). Without it, the invoice risked rejection.
Compliance here wasn't optional: it was a prerequisite for a mandatory certification that determined whether the product could keep serving this segment of practitioners at all. A mistyped or misdated renewal put the practitioner at direct financial risk. And the vocabulary didn't line up — practitioners talk about “extending,” “continuing,” or “starting a new plan,” while the regulatory classification is closed and fixed. Part of the design work, done with the team's UX writer, was finding the practitioner's words while still producing the data the regulator expected.
Writing in the practitioner's words rather than the regulation's was an accessibility decision as much as a wording one: the compliant path had to be usable without learning the classification.
The challenge
How do you replace a documented workaround that people use every day with a compliant path that's understandable enough — and simple enough — to actually change an established habit?
2 · Approach
The most useful insight
The workaround that already existed. Before this project, practitioners had already solved the problem themselves — with a documented twelve-step workaround: creating a fictitious prescriber with a dedicated identifier, purely to record the renewal. That workaround was documented, shared peer to peer, and used daily. That's a strong signal: when users have already built their own solution, the need doesn't need proving anymore. The question becomes whether your solution is enough better that they'll actually switch.
What we built
A renewal flow integrated into the treatment plan, in four steps — open the patient's treatment plan, trigger “Renew,” choose the renewal type (worded around practitioner language, not regulatory terms), confirm with a summary. The system generates the compliance data automatically and transmits the digital supporting document at the point of electronic submission. Fields are pre-filled from the original prescription. The historical workaround was retired. Rollout went by successive cohorts, to manage support load — a direct application of the lesson from the payment-guarantee case.
3 · Outcome
Flow shipped by cohort, and the historical workaround retired. Adoption levels, rejection reduction, and completion time remain targets rather than documented results at this stage.
4 · Learnings
Habit has inertia. The main risk on this project wasn't technical or regulatory, it was behavioral: getting practitioners who'd relied on a workaround for months to switch. That means designing not just the flow, but the moment and manner in which people discover it.